ultraqa

Automated QA cycles resolve failing Node and front-end builds by iteratively running tests, lints, and type checks, diagnosing failures, and applying fixes until quality goals are met. It coordinates execution, architect analysis, and executor fixes with state persistence across up to five cycles.

Autonomous build verification runs iterative QA cycles that execute tests, diagnose failures with an architect agent, and apply executor fixes. It maintains failure tracking and state persistence in a local JSON file, repeating until tests pass or a safe exit is triggered.

Automated cycle management supports custom success patterns alongside standard tests, builds, lints, and type checks. You activate it by describing your quality objective, allowing it to parse the goal and begin cycling through execution and fixes for your repository.

Automated QA cycle management stops safely if builds keep failing by enforcing cycle limits up to five iterations and explicit cancellation handling. It deletes stale state files upon completion or cancellation to ensure clear cycle output and prevent infinite loops.

State tracking for iterative test suite diagnosis maintains cycle progress inside a local JSON file, ensuring failure logs and architect diagnostics persist across iterations. This observability prevents duplicate work and enables executor-driven fixes to build upon previous cycle results.