Flake or Fault
Proves why tests fail — then fixes the real cause, not the symptom.
About
A skill that diagnoses failing tests by rerunning them to distinguish four causes: broken by your change, already broken on main, flaky, or order-dependent.
It then fixes the actual root cause instead of papering over it with retries or sleeps. The trade-off is time: the rerun-based proof is slower than guessing, which is the point.
What it can do
- Failure classification (change vs main vs flaky vs order)
- Rerun-based flakiness proof
- Root-cause fixing
- No retry-or-sleep workarounds
- Regression-safe repairs
Setup
- Add the marketplace to Claude Code.
- Install the flake-or-fault plugin.
- Point it at the failing tests and let it rerun and classify before fixing.
Prerequisites
A code repository with a repeatable test runner.
Use cases
Works with
Trust score — 5/10
Scored 5/10 (metrics tier, metrics-v1). Evidence: 1 star on the catalog repo, very early; rerun-based proof is slower than guessing by design; fixes root causes, not symptoms. License: MIT.
★ 1 GitHub stars
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 3.0/10
- Maintenance · recency, releases, issue load
- 5.0/10
- Openness · license + docs
- 8.0/10
- Manual review · sanity read + red-flag check
- 6.0/10
- · 1 star on the catalog repo, very early
- · rerun-based proof is slower than guessing by design
- · fixes root causes, not symptoms
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.