dataset
stringclasses 2
values | data_source
stringclasses 1
value | prompt
listlengths 2
2
| ability
stringclasses 1
value | reward_model
dict | extra_info
dict |
|---|---|---|---|---|---|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What year saw the creation of the region where the county of Hertfordshire is located? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1994",
"style": "rule"
}
|
{
"answer": "1994",
"evidence_list": [],
"index": 1,
"need_tools_kwargs": true,
"question": "What year saw the creation of the region where the county of Hertfordshire is located?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1994"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1994"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When was the abolishment of the studio that distributed The Game? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1999",
"style": "rule"
}
|
{
"answer": "1999",
"evidence_list": [],
"index": 2,
"need_tools_kwargs": true,
"question": "When was the abolishment of the studio that distributed The Game?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1999"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1999"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Where did the Baldevins bryllup director die? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Copenhagen",
"style": "rule"
}
|
{
"answer": "Copenhagen",
"evidence_list": [],
"index": 7,
"need_tools_kwargs": true,
"question": "Where did the Baldevins bryllup director die?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Copenhagen"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Copenhagen"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Which major Russian city borders the body of water in which Saaremaa is located? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Saint Petersburg",
"style": "rule"
}
|
{
"answer": "Saint Petersburg",
"evidence_list": [],
"index": 9,
"need_tools_kwargs": true,
"question": "Which major Russian city borders the body of water in which Saaremaa is located?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Saint Petersburg"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Saint Petersburg"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When was the employer of John J. Collins established? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1822",
"style": "rule"
}
|
{
"answer": "1822",
"evidence_list": [],
"index": 10,
"need_tools_kwargs": true,
"question": "When was the employer of John J. Collins established?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1822"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1822"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In what year was the author of The Insider's Guide to the Colleges established? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1878",
"style": "rule"
}
|
{
"answer": "1878",
"evidence_list": [],
"index": 14,
"need_tools_kwargs": true,
"question": "In what year was the author of The Insider's Guide to the Colleges established?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1878"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1878"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Where did the director who provided the lyrics to A Time for Singing attend school? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Northwestern's School of Communication",
"style": "rule"
}
|
{
"answer": "Northwestern's School of Communication",
"evidence_list": [],
"index": 17,
"need_tools_kwargs": true,
"question": "Where did the director who provided the lyrics to A Time for Singing attend school?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Northwestern's School of Communication"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Northwestern's School of Communication"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What languages are spoken, written or signed by the person the Schiff test is named after? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Italian",
"style": "rule"
}
|
{
"answer": "Italian",
"evidence_list": [],
"index": 32,
"need_tools_kwargs": true,
"question": "What languages are spoken, written or signed by the person the Schiff test is named after?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Italian"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Italian"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What is the language for the person who translated Plato's work to Latin? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Italian",
"style": "rule"
}
|
{
"answer": "Italian",
"evidence_list": [],
"index": 36,
"need_tools_kwargs": true,
"question": "What is the language for the person who translated Plato's work to Latin?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Italian"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Italian"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who was the president of the body that petitioned the king of England to recognize the rights of colonists when the declaration of independence was adopted? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "President John Hancock",
"style": "rule"
}
|
{
"answer": "President John Hancock",
"evidence_list": [],
"index": 38,
"need_tools_kwargs": true,
"question": "Who was the president of the body that petitioned the king of England to recognize the rights of colonists when the declaration of independence was adopted?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "President John Hancock"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "President John Hancock"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In what year did France host the finals of the competition featuring Samoa for going on 30 years? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "2007",
"style": "rule"
}
|
{
"answer": "2007",
"evidence_list": [],
"index": 41,
"need_tools_kwargs": true,
"question": "In what year did France host the finals of the competition featuring Samoa for going on 30 years?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "2007"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "2007"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In which year did the publisher of Roald Dahl's Guide to Railway Safety cease to exist? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "2001",
"style": "rule"
}
|
{
"answer": "2001",
"evidence_list": [],
"index": 42,
"need_tools_kwargs": true,
"question": "In which year did the publisher of Roald Dahl's Guide to Railway Safety cease to exist?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "2001"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "2001"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When was the inception of the company that released Poptropica? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1998",
"style": "rule"
}
|
{
"answer": "1998",
"evidence_list": [],
"index": 43,
"need_tools_kwargs": true,
"question": "When was the inception of the company that released Poptropica?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1998"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1998"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What type of community is the municipality where Norbert Pfretzschner died, an instance of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "comune",
"style": "rule"
}
|
{
"answer": "comune",
"evidence_list": [],
"index": 49,
"need_tools_kwargs": true,
"question": "What type of community is the municipality where Norbert Pfretzschner died, an instance of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "comune"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "comune"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What conflict involved the most famous Confederacy army? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "American Civil War",
"style": "rule"
}
|
{
"answer": "American Civil War",
"evidence_list": [],
"index": 53,
"need_tools_kwargs": true,
"question": "What conflict involved the most famous Confederacy army?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "American Civil War"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "American Civil War"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What is the name of the airport in the city where the torch event was held in Vietnam? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Tân Sơn Nhất International Airport",
"style": "rule"
}
|
{
"answer": "Tân Sơn Nhất International Airport",
"evidence_list": [],
"index": 54,
"need_tools_kwargs": true,
"question": "What is the name of the airport in the city where the torch event was held in Vietnam?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Tân Sơn Nhất International Airport"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Tân Sơn Nhất International Airport"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What year was the end of the manufacturer of the White armored car? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1980",
"style": "rule"
}
|
{
"answer": "1980",
"evidence_list": [],
"index": 57,
"need_tools_kwargs": true,
"question": "What year was the end of the manufacturer of the White armored car?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1980"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1980"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What university is Anthony Marten's college part of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "college of the University of Cambridge",
"style": "rule"
}
|
{
"answer": "college of the University of Cambridge",
"evidence_list": [],
"index": 59,
"need_tools_kwargs": true,
"question": "What university is Anthony Marten's college part of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "college of the University of Cambridge"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "college of the University of Cambridge"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What university is the college attended by Roland Moyle part of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "college of the University of Cambridge",
"style": "rule"
}
|
{
"answer": "college of the University of Cambridge",
"evidence_list": [],
"index": 61,
"need_tools_kwargs": true,
"question": "What university is the college attended by Roland Moyle part of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "college of the University of Cambridge"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "college of the University of Cambridge"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What language is used by the director of Thenum Vayambum? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Hindi",
"style": "rule"
}
|
{
"answer": "Hindi",
"evidence_list": [],
"index": 72,
"need_tools_kwargs": true,
"question": "What language is used by the director of Thenum Vayambum?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Hindi"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Hindi"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What is the family of the current leader of Saudi Arabia? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Al Saud",
"style": "rule"
}
|
{
"answer": "Al Saud",
"evidence_list": [],
"index": 76,
"need_tools_kwargs": true,
"question": "What is the family of the current leader of Saudi Arabia?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Al Saud"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Al Saud"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What was the language of the designer of Sacro Cuore di Cristo Re? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Italian",
"style": "rule"
}
|
{
"answer": "Italian",
"evidence_list": [],
"index": 82,
"need_tools_kwargs": true,
"question": "What was the language of the designer of Sacro Cuore di Cristo Re?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Italian"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Italian"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who was President of the U.S. following the event which was preceded by Joseph Schumpeter and Norman Angell's most prolific writing ? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Thomas Woodrow Wilson",
"style": "rule"
}
|
{
"answer": "Thomas Woodrow Wilson",
"evidence_list": [],
"index": 84,
"need_tools_kwargs": true,
"question": "Who was President of the U.S. following the event which was preceded by Joseph Schumpeter and Norman Angell's most prolific writing ?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Thomas Woodrow Wilson"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Thomas Woodrow Wilson"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did Tito first visit the person who laid claim to Eritrea and Italian Somaliland in a letter to Franklin D. Roosevelt? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1954",
"style": "rule"
}
|
{
"answer": "1954",
"evidence_list": [],
"index": 85,
"need_tools_kwargs": true,
"question": "When did Tito first visit the person who laid claim to Eritrea and Italian Somaliland in a letter to Franklin D. Roosevelt?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1954"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1954"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who is the child of Liszt's mistress? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Cosima Wagner",
"style": "rule"
}
|
{
"answer": "Cosima Wagner",
"evidence_list": [],
"index": 89,
"need_tools_kwargs": true,
"question": "Who is the child of Liszt's mistress?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Cosima Wagner"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Cosima Wagner"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When was the record label for Lovely Creatures created? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "2003",
"style": "rule"
}
|
{
"answer": "2003",
"evidence_list": [],
"index": 91,
"need_tools_kwargs": true,
"question": "When was the record label for Lovely Creatures created?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "2003"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "2003"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the institute that gives out the Elliott Cresson Medal award open? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1824",
"style": "rule"
}
|
{
"answer": "1824",
"evidence_list": [],
"index": 93,
"need_tools_kwargs": true,
"question": "When did the institute that gives out the Elliott Cresson Medal award open?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1824"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1824"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What year did the company that built K200 KIFV dissolve? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1999",
"style": "rule"
}
|
{
"answer": "1999",
"evidence_list": [],
"index": 94,
"need_tools_kwargs": true,
"question": "What year did the company that built K200 KIFV dissolve?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1999"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1999"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What family does the mother of Louise of Great Britain belong to? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "House of Hohenzollern",
"style": "rule"
}
|
{
"answer": "House of Hohenzollern",
"evidence_list": [],
"index": 103,
"need_tools_kwargs": true,
"question": "What family does the mother of Louise of Great Britain belong to?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "House of Hohenzollern"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "House of Hohenzollern"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Which family is the mother of Anne, Princess Royal and Princess of Orange's a member of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "House of Hohenzollern",
"style": "rule"
}
|
{
"answer": "House of Hohenzollern",
"evidence_list": [],
"index": 110,
"need_tools_kwargs": true,
"question": "Which family is the mother of Anne, Princess Royal and Princess of Orange's a member of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "House of Hohenzollern"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "House of Hohenzollern"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who was the first president from the party following the Anti-Administration Party? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Jefferson",
"style": "rule"
}
|
{
"answer": "Jefferson",
"evidence_list": [],
"index": 111,
"need_tools_kwargs": true,
"question": "Who was the first president from the party following the Anti-Administration Party?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Jefferson"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Jefferson"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "During what years did Frédéric take lessons at the academy where Chopin's father once had a teaching position? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1823 to 1826",
"style": "rule"
}
|
{
"answer": "1823 to 1826",
"evidence_list": [],
"index": 113,
"need_tools_kwargs": true,
"question": "During what years did Frédéric take lessons at the academy where Chopin's father once had a teaching position?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1823 to 1826"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1823 to 1826"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In what year did the agency that maintains the Presidential Library system become independent? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1985",
"style": "rule"
}
|
{
"answer": "1985",
"evidence_list": [],
"index": 115,
"need_tools_kwargs": true,
"question": "In what year did the agency that maintains the Presidential Library system become independent?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1985"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1985"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In what year did the South Korean car manufacturer that purchased the factor in 1995 dissolve? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1999",
"style": "rule"
}
|
{
"answer": "1999",
"evidence_list": [],
"index": 116,
"need_tools_kwargs": true,
"question": "In what year did the South Korean car manufacturer that purchased the factor in 1995 dissolve?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1999"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1999"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What was a notable work of the Chet Holifield Federal Building designer? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Transamerica Pyramid",
"style": "rule"
}
|
{
"answer": "Transamerica Pyramid",
"evidence_list": [],
"index": 117,
"need_tools_kwargs": true,
"question": "What was a notable work of the Chet Holifield Federal Building designer?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Transamerica Pyramid"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Transamerica Pyramid"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What establishment today contains what was known as the place Chopin's father got a teaching position during that time? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Warsaw University",
"style": "rule"
}
|
{
"answer": "Warsaw University",
"evidence_list": [],
"index": 118,
"need_tools_kwargs": true,
"question": "What establishment today contains what was known as the place Chopin's father got a teaching position during that time?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Warsaw University"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Warsaw University"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the sports team that Junior Félix played for win their last World Series? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1993",
"style": "rule"
}
|
{
"answer": "1993",
"evidence_list": [],
"index": 120,
"need_tools_kwargs": true,
"question": "When did the sports team that Junior Félix played for win their last World Series?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1993"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1993"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who was the first president from the newly evolved political party Jeremiah Morrow was a member of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Jefferson",
"style": "rule"
}
|
{
"answer": "Jefferson",
"evidence_list": [],
"index": 121,
"need_tools_kwargs": true,
"question": "Who was the first president from the newly evolved political party Jeremiah Morrow was a member of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Jefferson"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Jefferson"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Where was the artist of Honeyburst born? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Copenhagen",
"style": "rule"
}
|
{
"answer": "Copenhagen",
"evidence_list": [],
"index": 122,
"need_tools_kwargs": true,
"question": "Where was the artist of Honeyburst born?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Copenhagen"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Copenhagen"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What company published the cartoon that included Barnyard Dawg? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Warner Bros.",
"style": "rule"
}
|
{
"answer": "Warner Bros.",
"evidence_list": [],
"index": 124,
"need_tools_kwargs": true,
"question": "What company published the cartoon that included Barnyard Dawg?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Warner Bros."
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Warner Bros."
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who produced the series Acme Animation Factory is part of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Warner Bros.",
"style": "rule"
}
|
{
"answer": "Warner Bros.",
"evidence_list": [],
"index": 127,
"need_tools_kwargs": true,
"question": "Who produced the series Acme Animation Factory is part of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Warner Bros."
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Warner Bros."
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did person who supervised the design and implementation of the iPod user interface die? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "2011",
"style": "rule"
}
|
{
"answer": "2011",
"evidence_list": [],
"index": 128,
"need_tools_kwargs": true,
"question": "When did person who supervised the design and implementation of the iPod user interface die?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "2011"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "2011"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "The city where WNZK is located is in what county? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Wayne County",
"style": "rule"
}
|
{
"answer": "Wayne County",
"evidence_list": [],
"index": 130,
"need_tools_kwargs": true,
"question": "The city where WNZK is located is in what county?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Wayne County"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Wayne County"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In which county is the city in which KZOI is licensed? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Dakota County",
"style": "rule"
}
|
{
"answer": "Dakota County",
"evidence_list": [],
"index": 133,
"need_tools_kwargs": true,
"question": "In which county is the city in which KZOI is licensed?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Dakota County"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Dakota County"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In which year did the company that made SS.11 end? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1970",
"style": "rule"
}
|
{
"answer": "1970",
"evidence_list": [],
"index": 139,
"need_tools_kwargs": true,
"question": "In which year did the company that made SS.11 end?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1970"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1970"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the company that made the AA.20 end by merging? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1970",
"style": "rule"
}
|
{
"answer": "1970",
"evidence_list": [],
"index": 141,
"need_tools_kwargs": true,
"question": "When did the company that made the AA.20 end by merging?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1970"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1970"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In which year did the SS.11 manufacturer end? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1970",
"style": "rule"
}
|
{
"answer": "1970",
"evidence_list": [],
"index": 142,
"need_tools_kwargs": true,
"question": "In which year did the SS.11 manufacturer end?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1970"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1970"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What year was the end of the manufacturer of the Hudson Wasp? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1954",
"style": "rule"
}
|
{
"answer": "1954",
"evidence_list": [],
"index": 143,
"need_tools_kwargs": true,
"question": "What year was the end of the manufacturer of the Hudson Wasp?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1954"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1954"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What was the last year of the company that made the Hudson Commodore? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1954",
"style": "rule"
}
|
{
"answer": "1954",
"evidence_list": [],
"index": 144,
"need_tools_kwargs": true,
"question": "What was the last year of the company that made the Hudson Commodore?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1954"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1954"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who disarmed the combined body of all Estonia military? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "the Red Army",
"style": "rule"
}
|
{
"answer": "the Red Army",
"evidence_list": [],
"index": 149,
"need_tools_kwargs": true,
"question": "Who disarmed the combined body of all Estonia military?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "the Red Army"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "the Red Army"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Which company owns the company through which the TCM Vault Collection was released? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Warner Bros.",
"style": "rule"
}
|
{
"answer": "Warner Bros.",
"evidence_list": [],
"index": 152,
"need_tools_kwargs": true,
"question": "Which company owns the company through which the TCM Vault Collection was released?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Warner Bros."
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Warner Bros."
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "The writer of The Night They Drove Old Dixie Down falls under which record label? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Capitol Records",
"style": "rule"
}
|
{
"answer": "Capitol Records",
"evidence_list": [],
"index": 156,
"need_tools_kwargs": true,
"question": "The writer of The Night They Drove Old Dixie Down falls under which record label?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Capitol Records"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Capitol Records"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What record label is the musician who wrote the day they drove old dixie down affiliated? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Capitol Records",
"style": "rule"
}
|
{
"answer": "Capitol Records",
"evidence_list": [],
"index": 157,
"need_tools_kwargs": true,
"question": "What record label is the musician who wrote the day they drove old dixie down affiliated?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Capitol Records"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Capitol Records"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the military which introduced the tank to the battlefields of WWI start using the SA80? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1987",
"style": "rule"
}
|
{
"answer": "1987",
"evidence_list": [],
"index": 158,
"need_tools_kwargs": true,
"question": "When did the military which introduced the tank to the battlefields of WWI start using the SA80?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1987"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1987"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In what county is WGLB located? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Waukesha County",
"style": "rule"
}
|
{
"answer": "Waukesha County",
"evidence_list": [],
"index": 160,
"need_tools_kwargs": true,
"question": "In what county is WGLB located?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Waukesha County"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Waukesha County"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the army that introduced tanks to battlefields in WWI start wearing khaki? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "early 1900s",
"style": "rule"
}
|
{
"answer": "early 1900s",
"evidence_list": [],
"index": 161,
"need_tools_kwargs": true,
"question": "When did the army that introduced tanks to battlefields in WWI start wearing khaki?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "early 1900s"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "early 1900s"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In which county can you find WWGC? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Marshall County",
"style": "rule"
}
|
{
"answer": "Marshall County",
"evidence_list": [],
"index": 163,
"need_tools_kwargs": true,
"question": "In which county can you find WWGC?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Marshall County"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Marshall County"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What language was used by the cast member of Black Emanuelle? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Italian",
"style": "rule"
}
|
{
"answer": "Italian",
"evidence_list": [],
"index": 166,
"need_tools_kwargs": true,
"question": "What language was used by the cast member of Black Emanuelle?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Italian"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Italian"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "How many times did the winner of the 2018 WWE Elimination Chamber win the WWE championship? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "three",
"style": "rule"
}
|
{
"answer": "three",
"evidence_list": [],
"index": 172,
"need_tools_kwargs": true,
"question": "How many times did the winner of the 2018 WWE Elimination Chamber win the WWE championship?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "three"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "three"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "what language is spoken by the actress who decided not to participate in April 15? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Hindi",
"style": "rule"
}
|
{
"answer": "Hindi",
"evidence_list": [],
"index": 173,
"need_tools_kwargs": true,
"question": "what language is spoken by the actress who decided not to participate in April 15?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Hindi"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Hindi"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "How many times did the winner of the 7 man elimination chamber match win the WWE championship? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "three",
"style": "rule"
}
|
{
"answer": "three",
"evidence_list": [],
"index": 174,
"need_tools_kwargs": true,
"question": "How many times did the winner of the 7 man elimination chamber match win the WWE championship?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "three"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "three"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "How many times has the wrestler who won the WWE Elimination Chamber in 2018, won the WWE Championship? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "three",
"style": "rule"
}
|
{
"answer": "three",
"evidence_list": [],
"index": 175,
"need_tools_kwargs": true,
"question": "How many times has the wrestler who won the WWE Elimination Chamber in 2018, won the WWE Championship?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "three"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "three"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What language is used by the cast member of Emanuelle and the White Slave Trade? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Italian",
"style": "rule"
}
|
{
"answer": "Italian",
"evidence_list": [],
"index": 176,
"need_tools_kwargs": true,
"question": "What language is used by the cast member of Emanuelle and the White Slave Trade?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Italian"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Italian"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When was the record label of Factory Floor established? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "2001",
"style": "rule"
}
|
{
"answer": "2001",
"evidence_list": [],
"index": 185,
"need_tools_kwargs": true,
"question": "When was the record label of Factory Floor established?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "2001"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "2001"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Which title is used for the leader of the organization that has always opposed Freemasonry? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "the Bishop of Rome, who is known as the Pope",
"style": "rule"
}
|
{
"answer": "the Bishop of Rome, who is known as the Pope",
"evidence_list": [],
"index": 187,
"need_tools_kwargs": true,
"question": "Which title is used for the leader of the organization that has always opposed Freemasonry?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "the Bishop of Rome, who is known as the Pope"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "the Bishop of Rome, who is known as the Pope"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What marked the end of Titus Lucretius Tricipitinus's country? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "establishment of the Roman Empire.",
"style": "rule"
}
|
{
"answer": "establishment of the Roman Empire.",
"evidence_list": [],
"index": 206,
"need_tools_kwargs": true,
"question": "What marked the end of Titus Lucretius Tricipitinus's country?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "establishment of the Roman Empire."
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "establishment of the Roman Empire."
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the person Victoria recalled to office following her disappointment with Derby in June of 1859 die? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1865",
"style": "rule"
}
|
{
"answer": "1865",
"evidence_list": [],
"index": 208,
"need_tools_kwargs": true,
"question": "When did the person Victoria recalled to office following her disappointment with Derby in June of 1859 die?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1865"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1865"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Where did the patriots live during the war following which Boston became one of the wealthiest international ports? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "the Thirteen Colonies",
"style": "rule"
}
|
{
"answer": "the Thirteen Colonies",
"evidence_list": [],
"index": 211,
"need_tools_kwargs": true,
"question": "Where did the patriots live during the war following which Boston became one of the wealthiest international ports?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "the Thirteen Colonies"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "the Thirteen Colonies"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the UK vote to leave the organization which the Swiss have brought their economic practices into conformity with? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "23 June 2016",
"style": "rule"
}
|
{
"answer": "23 June 2016",
"evidence_list": [],
"index": 218,
"need_tools_kwargs": true,
"question": "When did the UK vote to leave the organization which the Swiss have brought their economic practices into conformity with?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "23 June 2016"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "23 June 2016"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the person who resigned after the scandal of Orsini's bomb being made in England die? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1865",
"style": "rule"
}
|
{
"answer": "1865",
"evidence_list": [],
"index": 225,
"need_tools_kwargs": true,
"question": "When did the person who resigned after the scandal of Orsini's bomb being made in England die?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1865"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1865"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the person who resigned their position following the scandal due to Orsini's bomb being made in England die? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1865",
"style": "rule"
}
|
{
"answer": "1865",
"evidence_list": [],
"index": 227,
"need_tools_kwargs": true,
"question": "When did the person who resigned their position following the scandal due to Orsini's bomb being made in England die?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1865"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1865"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Where did the person the National Gandhi Museum is named after spend their final 144 days? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Gandhi Smriti",
"style": "rule"
}
|
{
"answer": "Gandhi Smriti",
"evidence_list": [],
"index": 229,
"need_tools_kwargs": true,
"question": "Where did the person the National Gandhi Museum is named after spend their final 144 days?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Gandhi Smriti"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Gandhi Smriti"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Elizabeth Claypole's father started burying national figures in the abbey starting with whom? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Admiral Robert Blake",
"style": "rule"
}
|
{
"answer": "Admiral Robert Blake",
"evidence_list": [],
"index": 230,
"need_tools_kwargs": true,
"question": "Elizabeth Claypole's father started burying national figures in the abbey starting with whom?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Admiral Robert Blake"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Admiral Robert Blake"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Where did the leader of the Non-Cooperation movement in South India spend his final 144 days? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Gandhi Smriti",
"style": "rule"
}
|
{
"answer": "Gandhi Smriti",
"evidence_list": [],
"index": 233,
"need_tools_kwargs": true,
"question": "Where did the leader of the Non-Cooperation movement in South India spend his final 144 days?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Gandhi Smriti"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Gandhi Smriti"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who was the first national figure buried in the abbey by the father of Mary Cromwell, Countess Fauconberg? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Admiral Robert Blake",
"style": "rule"
}
|
{
"answer": "Admiral Robert Blake",
"evidence_list": [],
"index": 235,
"need_tools_kwargs": true,
"question": "Who was the first national figure buried in the abbey by the father of Mary Cromwell, Countess Fauconberg?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Admiral Robert Blake"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Admiral Robert Blake"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who was the first person buried in Westminster Abbey by the only non royal ruler of Britain? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Admiral Robert Blake",
"style": "rule"
}
|
{
"answer": "Admiral Robert Blake",
"evidence_list": [],
"index": 236,
"need_tools_kwargs": true,
"question": "Who was the first person buried in Westminster Abbey by the only non royal ruler of Britain?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Admiral Robert Blake"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Admiral Robert Blake"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "The ruler who granted the English East India Company a charter to govern Saint Helena began burying national figures in the abbey, starting with whom? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Admiral Robert Blake",
"style": "rule"
}
|
{
"answer": "Admiral Robert Blake",
"evidence_list": [],
"index": 237,
"need_tools_kwargs": true,
"question": "The ruler who granted the English East India Company a charter to govern Saint Helena began burying national figures in the abbey, starting with whom?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Admiral Robert Blake"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Admiral Robert Blake"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In what year was the band that performed Hope created? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "2001",
"style": "rule"
}
|
{
"answer": "2001",
"evidence_list": [],
"index": 242,
"need_tools_kwargs": true,
"question": "In what year was the band that performed Hope created?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "2001"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "2001"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In what year did the performers of Ohio Is for Lovers form? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "2001",
"style": "rule"
}
|
{
"answer": "2001",
"evidence_list": [],
"index": 246,
"need_tools_kwargs": true,
"question": "In what year did the performers of Ohio Is for Lovers form?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "2001"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "2001"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What group of schools is the university where Michael Berland studied a member of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Five Colleges",
"style": "rule"
}
|
{
"answer": "Five Colleges",
"evidence_list": [],
"index": 249,
"need_tools_kwargs": true,
"question": "What group of schools is the university where Michael Berland studied a member of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Five Colleges"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Five Colleges"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What type of school is the university that published the Owen Wister Review, an instance of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "land-grant university",
"style": "rule"
}
|
{
"answer": "land-grant university",
"evidence_list": [],
"index": 250,
"need_tools_kwargs": true,
"question": "What type of school is the university that published the Owen Wister Review, an instance of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "land-grant university"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "land-grant university"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What does the word mean that is a major part of Buddhism that rejects bardo? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "The School of the Elders",
"style": "rule"
}
|
{
"answer": "The School of the Elders",
"evidence_list": [],
"index": 255,
"need_tools_kwargs": true,
"question": "What does the word mean that is a major part of Buddhism that rejects bardo?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "The School of the Elders"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "The School of the Elders"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What year was the largest Presbyterian denomination brought to Lalrin Fela's birthplace? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1894",
"style": "rule"
}
|
{
"answer": "1894",
"evidence_list": [],
"index": 263,
"need_tools_kwargs": true,
"question": "What year was the largest Presbyterian denomination brought to Lalrin Fela's birthplace?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1894"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1894"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What year was the largest Presbyterian denomination brought to Pu Laldenga's birthplace? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1894",
"style": "rule"
}
|
{
"answer": "1894",
"evidence_list": [],
"index": 265,
"need_tools_kwargs": true,
"question": "What year was the largest Presbyterian denomination brought to Pu Laldenga's birthplace?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1894"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1894"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What is the meaning of the name for the field of work practiced at Wat Ananda Metyarama Thai Buddhist Temple? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "The School of the Elders",
"style": "rule"
}
|
{
"answer": "The School of the Elders",
"evidence_list": [],
"index": 267,
"need_tools_kwargs": true,
"question": "What is the meaning of the name for the field of work practiced at Wat Ananda Metyarama Thai Buddhist Temple?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "The School of the Elders"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "The School of the Elders"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What year was the largest Presbyterian denomination brought to the state where the Aizawl Monorail is located? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1894",
"style": "rule"
}
|
{
"answer": "1894",
"evidence_list": [],
"index": 269,
"need_tools_kwargs": true,
"question": "What year was the largest Presbyterian denomination brought to the state where the Aizawl Monorail is located?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1894"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1894"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who founded the company that made Vacamatic? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Walter Percy Chrysler",
"style": "rule"
}
|
{
"answer": "Walter Percy Chrysler",
"evidence_list": [],
"index": 271,
"need_tools_kwargs": true,
"question": "Who founded the company that made Vacamatic?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Walter Percy Chrysler"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Walter Percy Chrysler"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who founded the company that made the Plymouth Prowler? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Walter Percy Chrysler",
"style": "rule"
}
|
{
"answer": "Walter Percy Chrysler",
"evidence_list": [],
"index": 274,
"need_tools_kwargs": true,
"question": "Who founded the company that made the Plymouth Prowler?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Walter Percy Chrysler"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Walter Percy Chrysler"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What is the city where Giovanni Buonconsiglio was born an instance of? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "comune",
"style": "rule"
}
|
{
"answer": "comune",
"evidence_list": [],
"index": 276,
"need_tools_kwargs": true,
"question": "What is the city where Giovanni Buonconsiglio was born an instance of?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "comune"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "comune"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who was the spouse of the artist who created Circe Offering the Cup to Ulysses? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Esther Kenworthy Waterhouse",
"style": "rule"
}
|
{
"answer": "Esther Kenworthy Waterhouse",
"evidence_list": [],
"index": 280,
"need_tools_kwargs": true,
"question": "Who was the spouse of the artist who created Circe Offering the Cup to Ulysses?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Esther Kenworthy Waterhouse"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Esther Kenworthy Waterhouse"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In which county is the town in which WFQY is licensed? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Rankin County",
"style": "rule"
}
|
{
"answer": "Rankin County",
"evidence_list": [],
"index": 283,
"need_tools_kwargs": true,
"question": "In which county is the town in which WFQY is licensed?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Rankin County"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Rankin County"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When Anson led the Chicago Cubs, how many pennants did they win in the league that employed Peter McLaughlin? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "six",
"style": "rule"
}
|
{
"answer": "six",
"evidence_list": [],
"index": 290,
"need_tools_kwargs": true,
"question": "When Anson led the Chicago Cubs, how many pennants did they win in the league that employed Peter McLaughlin?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "six"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "six"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "While Anson led the Chicago Cubs, how many pennants did they win in the league that employed Bill Engeln? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "six",
"style": "rule"
}
|
{
"answer": "six",
"evidence_list": [],
"index": 291,
"need_tools_kwargs": true,
"question": "While Anson led the Chicago Cubs, how many pennants did they win in the league that employed Bill Engeln?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "six"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "six"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "While Anson led the Chicago cubs, how many pennants of the league having the Montreal Expos did they win? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "six",
"style": "rule"
}
|
{
"answer": "six",
"evidence_list": [],
"index": 293,
"need_tools_kwargs": true,
"question": "While Anson led the Chicago cubs, how many pennants of the league having the Montreal Expos did they win?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "six"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "six"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who is the mother of Kim Pyong-il's sibling? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Kim Jong-suk",
"style": "rule"
}
|
{
"answer": "Kim Jong-suk",
"evidence_list": [],
"index": 295,
"need_tools_kwargs": true,
"question": "Who is the mother of Kim Pyong-il's sibling?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Kim Jong-suk"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Kim Jong-suk"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "Who is the wife of Kim Jong-chul? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Kim Jong-suk",
"style": "rule"
}
|
{
"answer": "Kim Jong-suk",
"evidence_list": [],
"index": 301,
"need_tools_kwargs": true,
"question": "Who is the wife of Kim Jong-chul?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Kim Jong-suk"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Kim Jong-suk"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "When did the author of the The Cloud die? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "8 July 1822",
"style": "rule"
}
|
{
"answer": "8 July 1822",
"evidence_list": [],
"index": 305,
"need_tools_kwargs": true,
"question": "When did the author of the The Cloud die?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "8 July 1822"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "8 July 1822"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What position did the person who The Sun supported in 1974 hold? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "Prime Minister of the United Kingdom",
"style": "rule"
}
|
{
"answer": "Prime Minister of the United Kingdom",
"evidence_list": [],
"index": 306,
"need_tools_kwargs": true,
"question": "What position did the person who The Sun supported in 1974 hold?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "Prime Minister of the United Kingdom"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "Prime Minister of the United Kingdom"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "What year did the employer of William Sharp die? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1820",
"style": "rule"
}
|
{
"answer": "1820",
"evidence_list": [],
"index": 308,
"need_tools_kwargs": true,
"question": "What year did the employer of William Sharp die?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1820"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1820"
}
}
}
}
|
musique_train
|
musique_train
|
[
{
"content": "\nIn this environment you have access to a set of tools you can use to answer the user's question.\n\nYou only have access to the tools provided below. You can only use one tool per message, and will receive the result of that tool in the user's next response. You use tools step-by-step to accomplish a given task, with each tool-use informed by the result of the previous tool-use. Today is: 2025-07-16\n\n# Tool-Use Formatting Instructions\n\nTool-use is formatted using XML-style tags. The tool-use is enclosed in <use_mcp_tool></use_mcp_tool> and each parameter is similarly enclosed within its own set of tags.\n\nThe Model Context Protocol (MCP) connects to servers that provide additional tools and resources to extend your capabilities. You can use the server's tools via the `use_mcp_tool`.\n\nDescription:\nRequest to use a tool provided by a MCP server. Each MCP server can provide multiple tools with different capabilities. Tools have defined input schemas that specify required and optional parameters.\n\nParameters:\n- server_name: (required) The name of the MCP server providing the tool\n- tool_name: (required) The name of the tool to execute\n- arguments: (required) A JSON object containing the tool's input parameters, following the tool's input schema, quotes within string must be properly escaped, ensure it's valid JSON\n \nUsage:\n<use_mcp_tool>\n<server_name>server name here</server_name>\n<tool_name>tool name here</tool_name>\n<arguments>\n{\n \"param1\": \"value1\",\n \"param2\": \"value2 \"escaped string\"\"\n}\n</arguments>\n</use_mcp_tool>\n\nImportant Notes:\n- Tool-use must be placed **at the end** of your response, **top-level**, and not nested within other tags.\n- Always adhere to this format for the tool use to ensure proper parsing and execution.\n\nString and scalar parameters should be specified as is, while lists and objects should use JSON format. Note that spaces for string values are not stripped. The output is not expected to be valid XML and is parsed with regular expressions.\n\n\n\nHere are the functions available in JSONSchema format:\n\n## Server name: browsing_agent\n### Tool name: search_and_browse\nDescription: This tool is an agent that performs the subtask of searching and browsing the web for specific missing information and generating the desired answer. The subtask should be clearly defined, include relevant background, and focus on factual gaps. It does not perform vague or speculative subtasks. \nArgs: \n\tsubtask: the subtask to be performed. \nReturns: \n\tthe result of the subtask.\nInput JSON schema: {'properties': {'subtask': {'title': 'Subtask', 'type': 'string'}}, 'required': ['subtask'], 'title': 'search_and_browseArguments', 'type': 'object'}\n\n\n\n# General Objective\n\nYou accomplish a given task iteratively, breaking it down into clear steps and working through them methodically.\n\n## Task Strategy\n\n1. Analyze the user's request and set clear, achievable sub-goals. Prioritize these sub-goals in a logical order.\n2. Start with a concise, numbered, step-by-step plan outlining how you will solve the task before taking any action.\n3. Work through these sub-goals sequentially. After each step, adjust your plan as needed.\n4. Use tools strategically to accomplish each sub-goal.\n5. Revise earlier steps if new information emerges.\n\n## Tool-Use Guidelines\n\n1. Each step must involve a single tool call, unless the task is already solved.\n2. Before each tool call:\n - Summarize what is known.\n - Identify what is missing.\n - Choose the most relevant tool.\n - Verify all required parameters.\n3. All tool queries must include full context.\n4. Avoid vague queries. Each call should retrieve actionable information.\n5. Extract and summarize partial information if a tool result is incomplete.\n\n## Tool-Use Communication Rules\n\n1. Do not include tool results in your response.\n2. Do not present the final answer until the entire task is complete.\n3. Do not mention tool names.\n4. Do not engage in unnecessary back-and-forth.\n5. Do not use non-existent tools.\n6. Respond in the same language as the user's message.\n7. If the task does not require tool use, answer directly.\n\n\n# Agent Specific Objective\n\nYou are a task-solving agent that uses tools step-by-step to answer the user's question. Your goal is to provide complete, accurate and well-reasoned answers using additional tools.\n\n\n",
"role": "system"
},
{
"content": "In what year was the termination of the company that worked on USRA Light Pacific? You should follow the format instruction in the requestion strictly and wrap the final answer in \\boxed{}.",
"role": "user"
}
] |
qa
|
{
"ground_truth": "1920",
"style": "rule"
}
|
{
"answer": "1920",
"evidence_list": [],
"index": 310,
"need_tools_kwargs": true,
"question": "In what year was the termination of the company that worked on USRA Light Pacific?",
"question_type": "inference_query",
"split": "test",
"tools_kwargs": {
"google_search": {
"create_kwargs": {
"ground_truth": "1920"
}
},
"scrape": {
"create_kwargs": {
"ground_truth": "1920"
}
}
}
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.