eigent/backend/benchmark/dataset/1.json
bytecii f7bf29a40a
benchmark: update benchmark (#1207)
Co-authored-by: bytecii <bytecii@users.noreply.github.com>
Co-authored-by: Wendong-Fan <w3ndong.fan@gmail.com>
Co-authored-by: Wendong-Fan <133094783+Wendong-Fan@users.noreply.github.com>
2026-02-12 16:35:18 +08:00

22 lines
933 B
JSON

{
"metadata": {
"difficulty": "easy",
"description": "1) agent autonomously triggers search/browser to retrieve real data instead of hallucinating, 2) browser use with scrolling to locate specific modules, 3) instruction following for file creation with specific name and format.",
"tags": [
"browser",
"research",
"markdown",
"instruction-following",
"code-related"
]
},
"data": {
"name": "1",
"question": "Find what's new in Python 3.13 for the `warnings` and `multiprocessing` modules. Create a markdown file named 'python313_features.md' with each module name as a heading (#) and the exact text description from the official documentation as the content below each heading. Only make sure any code or script references are wrapped in backticks.",
"env": {}
},
"tests": {
"checker": ["benchmark/checker/1.py"],
"grader": ["benchmark/grader/1.py"]
}
}