aiw-verificationlisted

Governs how the agent confirms a change actually works before declaring a task complete. Use this skill at the end of every implementation pass and whenever the agent is about to say 'done', 'this works', 'tests pass', 'ready to commit', or 'ready to push'. Also use when interpreting runtime output, deciding whether to run the full system end-to-end, claiming a refactor preserves behaviour, a fix is fixed, a migration is complete, or a delete is safe. It forces the agent to justify every time that its checks cover the change and to name the unverified surface, breaking the 'all tests pass, system is broken' failure mode. It owns the required justification step before declaring done, the evidence hierarchy from static checks to end-to-end runs on real artifacts, the rules for when end-to-end execution is mandatory, reading runtime output as evidence not exit codes, modality-aware requirements referencing aiw-ground-truth, and the scoping step that names what was not checked.
philippe-ths/ai-coding-workflow · ★ 0 · AI & Automation · score 66

Install: claude install-skill philippe-ths/ai-coding-workflow

# Verification Read this file before declaring any task complete, before claiming "this works", before moving toward commit, push, or pull request, and whenever the agent is about to interpret tool output as success. ## Why This Skill Exists A passing test suite is not evidence the system works. It is evidence that the tests the agent chose to run passed, against the fixtures the agent chose to use, in the way the agent chose to interpret the output. Each of those choices is a place where the verification can be locally valid and globally wrong. The failure mode this skill exists to prevent: every check passes, the agent declares done, and the change breaks the system on first real use. The cause is almost always coverage — the checks did not actually exercise the path the change affected — rather than test correctness. The remedy is not more tests by default. It is forcing the agent to argue, explicitly and every time, that the checks it ran are sufficient evidence the change works. Verification is one of the three checks at the done gate. It asks whether the change is correct. aiw-validation asks whether the result is what was asked for, and work can satisfy this skill completely while delivering nothing the human wanted. aiw-housekeeping asks whether anything the task caused should now be removed or moved. Run all three. Classify the task using the modality decision procedure defined in aiw-ground-truth. Verification requirements depend on the modality. ## The Justi