aiw-testinglisted

Mechanical craft rules for writing tests well: choosing the right test level, consuming ground truth from aiw-ground-truth rather than fabricating it, organising tests for independence and clarity, and avoiding anti-patterns that produce green suites with no real authority. Use this skill whenever the agent is about to write or modify a test, choose a fixture, pick a test level (unit, integration, end-to-end), decide whether to mock a component, refactor test code, fill a coverage gap from aiw-verification, write a regression test, or when tempted to assert on internal calls or private state. It does not decide what counts as correct (aiw-ground-truth) or whether tests are sufficient evidence (aiw-verification). It owns test mechanics, test-level selection, modality-specific testing emphasis referencing aiw-ground-truth, the anti-patterns that produce misleading green bars, and the hygiene rules that keep a suite a reliable signal.
philippe-ths/ai-coding-workflow · ★ 0 · Testing & QA · score 66

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

# Testing Read this file when writing a new test, modifying an existing test, choosing a test level, picking a fixture, or deciding whether to mock a component. ## Why This Skill Exists Testing is the mechanical craft of constructing a check that the code produces the expected output. It is the easy part of the three-skill set: - **Ground truth** decides what counts as correct. The fixture, the expected output, the oracle. - **Testing** decides how to construct the check. - **Verification** decides whether the checks taken together are sufficient evidence the change works. This skill is narrow on purpose. Pull the oracle from aiw-ground-truth. Let aiw-verification judge whether the tests cover the change. Do not let testing absorb either question — the failure mode that motivates this whole set is one actor playing all the roles. Keeping testing's job mechanical preserves the separation. ## Consuming Ground Truth For inputs that represent real-system artifacts — real user data, real upstream outputs, real configurations — source the fixture from aiw-ground-truth. Do not fabricate the fixture inline and call it a test fixture; that is the closed-loop failure mode that motivates this entire set. For every test that depends on a ground-truth fixture: - Record the trust level of the fixture (the 1–5 hierarchy defined in aiw-ground-truth). - If the fixture is synthetic (trust level 5), mark the test visibly so aiw-verification treats its result as weaker evidence. - Do no