Autonomous Test-Repair Loop
$2.99OfficialUse when a repo's test suite is red and you must drive it to green by patching the code under test within a budget, without masking regressions.
What you get
- ✓11-step procedure
- ✓Runnable Python included
- ✓8-point quality checklist
- ✓10 pitfalls to avoid
- ✓Installs into 6 tools
- Version
- v1 →
- Last updated
- today
- Length
- 9 min read
- Requires
- Best with a strong model (Claude Opus 5)
Works in: Claude Code, Codex, Cline, opencode, OpenClaw, Hermes · Built for large codebases
Preview
When to use
Use when a repo has a runnable automated test suite that is red, and the goal is to drive it green by repairing the code under test — after a refactor, dependency bump, merge, rebase, or a half-finished feature. The core promise is a bounded loop: run, localize, patch, re-run, gated by a budget and a hard no-masking rule. Do NOT use to author new tests, to raise coverage, or when the failures are environmental (missing service, absent secret, offline network) rather than defects — fix the environment first, then re-enter here.
Inputs to gather
- Runner and exact commands: how to run the whole suite AND a single test by id (
pytest -q <nodeid>,jest -t "<name>", `vites
…
🔒 Buy once ($2.99) to unlock the full playbook, download it, and install it in every tool you use.