Revert proof

The idea

A new or changed test must fail when the fix is reverted.

The problem
A test that passes without the fix proves nothing.
Why it matters
Each fix has a test that shows it.
Where it stands
In production as of

Recorded walkthrough: A test must fail without the fix: the independent verifier. It shows the revert proof on two pull requests: the fix's new test failed without the fix, and a change that edited only a test was turned down.

  • Acceptance criteria as executable checks

    At plan time, each acceptance criterion maps to a test, script, query, HTTP check or screenshot check. Verification runs all of them in each round.

  • Independent verifier

    The platform checks an AI worker's pull request in its own container, out of the worker's reach, and approves it only when a new test fails without the fix.

    Recorded walkthrough

  • Completion truth gates

    New tests must run here or have a trace for each assertion. Claimed behaviour must be tested, and the same bug elsewhere fixed or ruled out. Planned changes must exist, steps tick only on evidence, and spend over budget needs a recorded decision.

Turn the numbers into shipped work.

Crewshal runs these choices for you: a persistent AI worker with memory and rules, on your Claude and Codex subscriptions.