A clean, modular SDK for building AI agents with OpenHands V1.
Python · active 2025-10-29 → 2026-08-09 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 285 days from 2025-10-29 to 2026-08-09. Pushes: 3,953 total, peak 74 in a day. Pull requests: 917 total, peak 25 in a day. Issues: 995 total, peak 57 in a day. Comments: 5,150 total, peak 112 in a day. Stars: 193 total, peak 11 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 |
|---|---|---|---|---|
| enyst | 2.1K | 693 | 120 | 699 |
| xingyaoww | 1.8K | 860 | 204 | 407 |
| all-hands-bot | 1.7K | 32 | 45 | 884 |
| openhands-ai[bot] | 1.5K | 0 | 0 | 1.3K |
| neubig | 943 | 474 | 105 | 181 |
| github-actions[bot] | 897 | 0 | 2 | 801 |
| simonrosenberg | 705 | 437 | 63 | 117 |
| VascoSch92 | 529 | 298 | 27 | 94 |
| juanmichelini | 391 | 186 | 44 | 81 |
| csmith49 | 355 | 218 | 41 | 57 |
| malhotra5 | 354 | 188 | 17 | 84 |
| jpshackelford | 302 | 140 | 12 | 87 |
| tofarr | 254 | 143 | 39 | 28 |
| ryanhoangt | 100 | 49 | 26 | 12 |
| rbren | 96 | 52 | 22 | 12 |
| ak684 | 87 | 54 | 7 | 11 |
| hieptl | 83 | 48 | 13 | 10 |
| blacksmith-sh[bot] | 79 | 0 | 0 | 79 |
| dependabot[bot] | 60 | 9 | 50 | 1 |
| aivong-openhands | 56 | 40 | 3 | 10 |
Recent activity
Latest issues, pull requests and releases
- Issue#4443all-hands-bot2026-08-09 22:51agent-server 1.41.0 PyInstaller bundle omits browser_use/js (rrweb recording)
- Issue#4268simonrosenberg2026-08-07 16:25Support concurrent-safe Codex auth refresh in Agent Canvas Docker
- Issue comment#4171simonrosenberg2026-08-07 16:22Phase 2: coordinate concurrent Codex credential refresh across ACP conversations
- Issue comment#4311hieptl2026-08-04 18:12feat: report accumulated LLM cost in the automation completion callback
- Issue comment#2509all-hands-bot2026-07-29 13:08feat(sdk): allow injecting a custom FileStore into ConversationState
- Issue comment#4284github-actions[bot]2026-07-28 00:59feat(llm): verify Kimi K3 and Claude Opus 5 with LiteLLM 1.93
- Issue comment#4153sideeffffect2026-07-27 13:32fix(llm): allow security_risk param on read-only tools like finish
- Issue#4231all-hands-bot2026-07-27 13:11[Bug]: test issue for triage automation — MCP server fails to connect on startup
- Issue comment#4196github-actions[bot]2026-07-22 16:52Release v1.37.0
- Issue comment#2772all-hands-bot2026-07-22 12:54feat(sdk): project-level .mcp.json discovery and merge
- Issue#2755VascoSch922026-07-20 10:40[Feature] Implement HookType.PROMPT — LLM-based hook evaluation
- Issue comment#3950github-actions[bot]2026-07-18 23:11[Tracker] Daily Examples Run Results
- Issue comment#3029all-hands-bot2026-07-16 12:58feat(sdk): introduce SnapshotReplayAgent for deterministic conversati…
- Issue comment#3991all-hands-bot2026-07-16 12:57fix(agent-server): survive webhook rate limiting and reap stale run tasks
- Issue#2053github-actions[bot]2026-07-16 02:19[Epic] Skills: Execution isolation, model routing & enhanced capabilities
- Issue comment#2078github-actions[bot]2026-07-13 22:55[Tracker] Daily Integration Runs
- Issue comment#3895all-hands-bot2026-07-11 12:34perf(agent-server): lazy-load persisted conversations at startup
- Issue#4079neubig2026-07-10 19:34fix(acp): tolerate missing token count for handled Codex commands
- Issue comment#4044github-actions[bot]2026-07-09 06:44chore(deps): bump lewagon/wait-on-check-action from 1.7.0 to 1.8.1
- Issue comment#4019enyst2026-07-08 01:07[sdk] ACP profiles inject workspace project skills that duplicate what the ACP CLI already ingests (AGENTS.md)
- Issue#2047github-actions[bot]2026-07-07 02:21feat(delegate): Non-blocking background subagent execution
- Issue comment#2697github-actions[bot]2026-07-07 02:21[Feature]: QuestionTool -- let the agent ask the user structured questions mid-task
- Pull request#4000aivong-openhands2026-07-05 21:51
- Pull request#3993neubig2026-07-05 12:58
- Issue#3291github-actions[bot]2026-07-04 02:20SwitchLLMTool: LLM can't detect external profile switches, leading to stale-state replies
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 193 stars here means stars gained during the window, not the repo's star count.