WindowsAgentArena Model Card
这是可索引项目报告证据页：它保留 WindowsAgentArena Model Card 的源材料入口、机制线索和限制提醒；正文仍需 reader/editor 与 academic public-copy review 后才能当作最终结论引用。
WindowsAgentArena Model Card
| Field | Value |
|---|---|
| Repository | microsoft/WindowsAgentArena |
| Category | Windows OS Agent Benchmark |
| Stars / forks snapshot | 861 / 95 |
| Language | Python |
| License | MIT |
| Raw capture | raw-github/microsoft_windowsagentarena.md |
| Updated by | hourly public metadata update, 2026-05-24 |
1. Role in Self Evolve
Microsoft Windows Agent Arena 是面向多模态 OS agents 的 Windows 评测平台，使用 Docker、Windows 11 VM/golden image 和 OpenAI/Azure OpenAI endpoint 执行大规模 benchmark。
2. Working Principle
Windows 11 VM/golden image -> multimodal OS agent -> scalable benchmark execution -> report metrics
3. Evidence Path
web GitHub page observed technical report citation, Docker/WSL/Python prerequisites, Windows 11 VM golden image workflow, and local deployment commands; 861 stars, 95 forks, 86 commits; shell GitHub API was blocked by DNS and gh auth was invalid, so this run marks freshness as web-page observed rather than API verified.
4. Teaching Use
Use this card to explain whether a repository is a runtime, benchmark, harness-evolution loop, memory/skill substrate, or resource index. The key reading path is: raw capture -> classification row -> public site card -> project report -> aggregate GitHub analysis.
5. Limits
当前未克隆源码，未运行 benchmark 或 SDK examples；star/fork/commit 快照来自公开 GitHub 页面文本或可见页面片段。