Evaluation harness for running multiple agent benchmarks against cuga-agent.
active 2026-05-26 → 2026-08-27 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 94 days from 2026-05-26 to 2026-08-27. Pushes: 72 total, peak 7 in a day. Pull requests: 2 total, peak 1 in a day. Issues: 5 total, peak 3 in a day. Comments: 5 total, peak 1 in a day. Stars: 0 total, peak 0 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 |
|---|---|---|---|---|
| haroldship | 61 | 55 | 0 | 2 |
| Iftach-Shoham | 13 | 12 | 1 | 0 |
| Sergey-Zeltyn | 7 | 4 | 1 | 2 |
| sami-marreed | 2 | 1 | 0 | 0 |
| coderabbitai[bot] | 1 | 0 | 0 | 1 |
Recent activity
Latest issues, pull requests and releases
- Issue#141haroldship2026-07-22 17:11fix(deps): pyasn1 0.6.3 fails pip-audit (PYSEC-2026-3455/3456/3457), fix in 0.6.4
- Issue comment#96Sergey-Zeltyn2026-07-02 08:11feat(m3): groundedness hardening — trim rule, gpt-4.1 judge backend, capability rollup, registry teardown fix
- Issue#24haroldship2026-06-25 23:38[Feature]: Evaluate CodeAct vs native function calling for CugaLite
- Issue#29haroldship2026-06-25 23:38[Feature]: Bring-your-own model (vLLM / local OpenAI-compatible) for all benchmarks
- Issue#17haroldship2026-06-25 23:38[Bug]: Port Hardcoding: load_eval_config sets APIS_URL, APPWORLD_ROOT, LITE_MODE_TOOL_THRESHOLD without documentation
- Pull request#80Iftach-Shoham2026-06-14 08:35
- Issue comment#77coderabbitai[bot]2026-06-14 08:32feat(reports): standardize Pass@1, cost stats, and group breakdowns in report.md
- Pull request#47Sergey-Zeltyn2026-06-07 10:03
- Issue comment#47Sergey-Zeltyn2026-06-07 10:02fix(analytics): skip temperature param for Claude 4.x models
- Issue comment#20haroldship2026-05-27 11:45Make AppWorld opt-in so uv sync works without it
- Issue#10sami-marreed2026-05-26 09:46M3: shared client-side rate limiter across parallel workers (Groq 429 cascade)
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 0 stars here means stars gained during the window, not the repo's star count.