Free tools we build and maintain for the developer community. Production-tested in our own projects, open source for everyone.
CodeAssay
Git forensics for AI-authored code quality
MIT License
Track the quality of AI-generated code across your repositories. Detects AI-authored commits from Claude, Copilot, and GPT, identifies rework, classifies root causes, and generates interactive dashboards. Works with any git repo — zero dependencies, fully offline.
Invariant
Prompt injection test harness for AI trust boundaries
TypeScript
Apache 2.0
Run canonical prompt injection benchmarks, author custom attack scenarios, and compare safety verdicts across LLM providers. Includes a FastAPI test harness and a React operator console. No API keys required — a mock provider covers end-to-end testing out of the box.
Build With Us
We open source the tools we build for our own AI development workflow. Contributions, issues, and feedback are welcome.