standup-bot: turns git history into a daily standup via local (Ollama) or free cloud (Groq) LLM.
Python · SQLite (WAL) · Rich
Fixed a real v0.2.3 bug (timeout passed as model param → requests hung forever) with a regression test. CI: ruff, mypy, bandit, pip-audit; SQL parameterized; secrets redacted from all output.
env-auditor: diffs env vars your code references against .env.example across six languages.
Python · zero runtime dependencies
Written defensively: skips long lines (ReDoS), never follows symlinks, --exclude can't escape the scan root. 141 tests, 85% coverage, three OSes × three Python versions, mypy --strict.
chess-review-bot: labels each PR file with chess Game Review vocabulary (Brilliant, Blunder...) via a deterministic rule table (no LLM call, reasoning in an ADR).
Python · GitHub Action + PyPI CLI
Shipped broken (Action doubled its own path, exit 2); caught by dogfooding on its own PRs, fixed in v0.1.1. Published via OIDC; secrets stripped from subprocess calls.
Also built position-evaluator for a weekend hackathon: MySQL + Flask + Gemini structured output; the LLM-reliability lessons made chess-review-bot deterministic instead.
Open source contributions
- Termstory: three fixes in three days, three subsystems; the circuit-breaker fix survived five review rounds (race condition + silent mutation trap).
- AynOps: root-caused four bugs behind its header analyzer vs. browser DevTools.
- composable-data-stack: hardened Docker image (pinned digest, non-root); closed a
.envleak into the build context. - thumper: fixed a session-hijack path + added ruff linting.
- odys: three fixes, incl. one where the issue's own suggested fix was mathematically wrong.
- Full list
Stats
