Refuses "flaky" as a diagnosis. Turns an intermittent failure into a reproduction by isolating it from the environment, naming the variable that controls it, forcing that variable, and measuring before and after with enough runs to count. Separates real defects from environment contention with evidence, and never retries to green or widens a threshold. Use when a test, performance check, or deploy step fails intermittently or passes on rerun.
The English summary is being prepared. Showing the original description.
See how to installReview the instructions the agent will receive before you install.
"Flaky" ends an investigation without finding anything, and a retry-to-green hides real defects. An intermittent failure is a hypothesis: some variable you haven't controlled decides the outcome. Find it, force it, and prove the fix with counts.
vibe-concurrent-test-safety)Record the failure exactly. Commit, environment, profile or device, the failing value, and the threshold. Keep the log.
Run it alone, cleanly. If it passes in isolation, treat it as contention until proven otherwise, and name the resource: CPU, a shared port, a shared database, disk, or network. Record machine load alongside every timing measurement.
List candidate variables and force each in turn. Common ones:
Reproduce at a stated rate before changing anything (for example "10 of 15 runs fail with fonts held back 600 ms"). No reproduction means no root cause yet, so keep looking or report it as unresolved.
Fix the cause, not the symptom. Then rerun the forced condition N times and report the new rate (for example "0 of 240 loads across 4 routes").
Disclose any mitigation. If you clip, retry, or quarantine anything, say what, why, and why it doesn't hide a defect. Never widen a threshold or budget to pass (vibe-anti-rationalization-check).
Symptom: [value vs threshold, where, how often] Isolated run: pass / fail · Load at failure: [value]
| Variable | Forced how | Failure rate |
|---|---|---|
| font arrival | held back 600 ms | 10/15 |
| CPU load | run alone | 0/15 |
Root cause: [mechanism, in one or two sentences] Fix: [change] · After: [0/N under the forced condition] Mitigations disclosed: [none / list with reasons]
No reviews yet. Tell others what you thought.
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
Validates completed work against defined acceptance criteria. Use after completing a task that has specific success criteria defined in issues, specs, or task descriptions.
The English summary is being prepared. Showing the original description.
Generates edge case, failure mode, and spec-driven test cases. Covers boundary values, nil inputs, concurrency, resource exhaustion, malformed data, and requirement-linked traceability tests. Use after happy-path tests exist and before claiming coverage is complete, when requirements lack tests, or before a security review.
The English summary is being prepared. Showing the original description.
Catches shortcuts and reward hacking — weakening or deleting tests to make them pass, hard-coding expected outputs, silently dropping requirements, or claiming work is done without running it. Use before declaring a task complete and whenever a test or requirement feels like it's in the way.
The English summary is being prepared. Showing the original description.
Safely integrates commits from parallel agent branches using sequential cherry-pick. Use after parallel work completes in isolated branches or worktrees.
The English summary is being prepared. Showing the original description.
Audits tests for concurrency safety — race conditions, shared mock state, cleanup ordering. Use when writing tests that involve goroutines, async operations, or shared mutable state.
The English summary is being prepared. Showing the original description.
Enforces tiered test coverage standards with three dimensions — line coverage by tier, spec-to-test traceability, and spec-to-code implementation mapping. Use before claiming code is complete.
The English summary is being prepared. Showing the original description.