Ta0p1 / ResolveLab
Fault-injected tool environments with side-effect grading for evaluating and post-training LLM agents
Why this job: topics: llm-agents, tool-calling, ai-agents
What it does
A tool-using agent can produce valid JSON, sound confident and still leave the system wrong: a write to the near-miss record, a retry that commits twice, "resolved" after a 502 that hid a lost acknowledgement. ResolveLab runs agents against a real HTTP service that injects those failures, and grades the episode from server state, the commit log and read versions, never from the agent's claim. The same environment provides a teacher whose labels are replay-verified, so it can also produce training data.
Get started
git clone https://github.com/Ta0p1/ResolveLab.git && cd ResolveLab
python -m pip install -e ".[dev]"
python examples/casedesk_quickstart.py - Maintenance
- Active
- Last commit yesterday
- Stars
- 104
- On GitHub
- Licence
- MIT
- Allows business use
- Downloads for
- None listed
- No release files
- Latest version
- v0.5.0
- Released 4 days ago
- Built with
- Python
- 92% of the code
Is it looked after?
- Contributors
- 1
- Last 90 days
- Issues
- None
- No issues filed
- Pull requests
- 18 merged
- 1 waiting
- Releases
- 2
- 2 in the last year
- Contributing guide (opens in new tab)
- Security policy (opens in new tab)
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| scaleapi / agentenv-framework | 420 | today | Apache-2.0 | No downloads |
| rzchen777-wq / MAC_Bench | 10 | today | MIT | No downloads |
More: languages, topics and project record
Languages
- Python 92%
- JavaScript 4%
- CSS 2%
- HTML 2%
Topics
- agent-evaluation
- ai-agents
- fastapi
- ollama
- sqlite
- tool-calling
- dagger
- fault-injection
- llm-agents
- post-training
Project record
- JobAgent frameworks
- StatusNew to our list
- Commits74
- Forks1
- Watching0
- CreatedOct 20, 2026
- First seen by usNov 7, 2026
- CheckedNov 9, 2026