这是一个让 AI 像人一样操作电脑（如 macOS、Windows、Linux、Android），可自动完成复杂的 GUI 操作任务的 AI Agent 框架。它采用 "通用体 + 专家体" 组合架构，支持主动分层规划。通过集成大模型（LLM）和视觉多模态模型，能够理解屏幕截图、界面结构等输入，并生成操作指令，实现自动点击、输入、窗口切换、搜索等操作
active 2024-10-11 → 2026-09-03 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 693 days from 2024-10-11 to 2026-09-03. Pushes: 263 total, peak 19 in a day. Pull requests: 89 total, peak 5 in a day. Issues: 128 total, peak 12 in a day. Comments: 277 total, peak 23 in a day. Stars: 7,098 total, peak 227 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| alckasoc | 329 | 180 | 29 | 67 |
| coderabbitai[bot] | 94 | 0 | 0 | 75 |
| eric-xw | 44 | 20 | 6 | 10 |
| saa1605 | 41 | 25 | 11 | 4 |
| eric-simu | 35 | 17 | 3 | 10 |
| GeorgeCarpenter | 28 | 0 | 1 | 17 |
| simularlyon | 19 | 6 | 3 | 8 |
| kylesimular | 16 | 13 | 0 | 3 |
| Richard-Simular | 16 | 3 | 0 | 10 |
| chenchenSimular | 10 | 0 | 2 | 5 |
| devin-ai-integration[bot] | 7 | 0 | 2 | 3 |
| tezansahu | 6 | 0 | 2 | 2 |
| pavel-kirienko | 5 | 0 | 0 | 2 |
| DaWe35 | 5 | 0 | 2 | 1 |
| rossamurphy | 5 | 0 | 1 | 2 |
| blake-lucas | 4 | 0 | 0 | 3 |
| yuhengcai1 | 4 | 0 | 1 | 2 |
| chalo2000 | 3 | 0 | 0 | 2 |
| kira-offgrid | 3 | 0 | 3 | 0 |
| AndryHTC | 3 | 0 | 0 | 1 |
Recent activity
Latest issues, pull requests and releases
- Pull request#193vikranth224462026-05-07 10:49
- Issue comment#178simularlyon2026-04-07 04:32Reduce grounding model call frequency
- Issue comment#186chenchenSimular2026-04-02 13:37Code execution tool broken
- Pull request#180haosenwang10182026-02-25 05:51
- Pull request#179simularhao2026-02-21 05:35
- Issue comment#177coderabbitai[bot]2026-02-11 02:38fix: improve coordinate precision with multi-sample voting
- Pull request#177Chase-Xuu2026-02-09 21:28
- Issue#176cppcoffee2026-02-02 01:36GLM-4.7 and GLM-Image have any demo?
- Issue comment#174coderabbitai[bot]2026-01-19 06:28chore: Update setup.py version to 0.3.2
- Issue comment#172Richard-Simular2026-01-19 05:17Fixed interval schedules
- Issue#172beezready2-bedW2026-01-16 20:55Fixed interval schedules
- Issue#171beezready2-bedW2026-01-16 09:45Fire Kirin
- Issue comment#170coderabbitai[bot]2026-01-14 21:22fix to follow style guideline for our linter
- Issue comment#162Richard-Simular2026-01-14 21:13tests: add final-status execution tests (done token and fail status)
- Pull request#156maxwellVisual2026-01-14 15:25
- Issue comment#167Richard-Simular2026-01-13 12:28Problem with bBoN
- Issue comment#61gysabc2026-01-13 09:08What knowledge / memory does the reported performance on OSWorld need?
- Issue comment#166coderabbitai[bot]2026-01-05 13:37Create Sock
- Pull request#166zgplaysgamer22-prog2026-01-05 13:37
- Pull request#165gator922026-01-02 17:48
- Issue comment#165coderabbitai[bot]2026-01-02 17:48Add y-offset for grounding coords
- Pull request#165gator922026-01-02 17:48
- Pull request#163majiayu0002025-12-30 11:54
- Issue comment#162coderabbitai[bot]2025-12-24 15:35tests: add final-status execution tests (done token and fail status)
- Issue comment#161Vasyl1982025-12-24 13:39feat: require user confirmation before executing agent-generated code
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 7,098 stars here means stars gained during the window, not the repo's star count.