← All work
Agent Ops Console
A console where reviewers flag agent mistakes; approved flags become regression tests.
32
pytest tests
A/B
on identical seeded traffic
What it does
Reviewers flag agent mistakes in a React and FastAPI console, and approved flags turn into regression tests. Releases are immutable and versioned, with rollback and A/B comparison on identical seeded traffic.
Stack
Python · FastAPI · SQLAlchemy · Pydantic · React · Vite · Recharts · Pytest
Honest notes
The code is private for now; I'm happy to walk through it.