Open-source product

qa-pilot

qa-pilot turns testing from an afterthought into a repeatable workflow that an AI coding agent can follow and verify.

Problem

AI coding tools can generate tests quickly, but generated coverage is often shallow, over-mocked, or unable to prove that it protects real behaviour.

A project onboarding flow that records the actual stack and test conventions
Separate workflows for test planning, implementation, exploratory testing, and suite audits
A mandatory proof that every new automated test can fail for the intended reason
Persistent project context so quality decisions survive across sessions

The plugin is publicly available through the Codex plugin ecosystem and is developed in the open on GitHub.

Describe the problem, current situation, and outcome you need. I will suggest a sensible first step.