Use when setting up or auditing a project's CI/CD pipeline together with local verification — ensures pnpm/npm verify (or equivalent) actually covers everything CI checks, adds pre-commit/pre-push git hooks that enforce this locally, and keeps trunk-based development honest (small commits, frequent pushes, CI checked per push not per batch). Load before building a new .woodpecker.yml/.github/workflows, before adding SonarQube/coverage, or when "it passed locally but failed in CI" comes up.

Find flaky and repeatedly failing tests from recent CI history and file a ticket for each one worth fixing. Use for a scheduled test-health check that catches test rot early, before a flaky test becomes one everyone just re-runs and ignores.

Use when tests need the app plus several real services (database, cache, queue) started together. Writes a compose.test.yaml with healthcheck-ordered startup, per-job project names for parallel CI, --wait and --exit-code-from gating, and clean teardown with volumes removed.

Use when non-production environments need masked production data. Classifies each field against GDPR, CCPA, NIST and HIPAA PII categories, picks a masking operator, decides pseudonymization vs anonymization, emits a Presidio, Faker and Synthea config and blocks promotion if a re-scan finds PII.

Use when a QA manager needs to know what the team can do today versus what its projects demand, before quarterly planning, a training budget decision or opening a role. Builds an evidence-backed skill matrix, classifies and ranks gaps and recommends train, hire or bring in an expert per gap.

Use before a QBR, board update or exec review when quality data exists (digests, KPIs, DORA metrics, escaped defects, OKR grades) but the story does not. Writes the executive narrative with the answer first and grouped support beneath it, using the Minto Pyramid and SCQA.

Use when a service worker needs lifecycle tests, users are stuck on an old version after deploy, or caching and offline behavior is untested. Tests each state change from installing to activated, waitUntil, skipWaiting and clients.claim, update() and a clean v1 to v2 upgrade with Playwright.

Use when a Rust project needs property-based tests, especially for parsers, serializers and functions with structured input. Sets up proptest: the proptest! macro, strategies for collections, regex strings and any, shrinking and the persisted regression file.

Use when a feature has no story, acceptance criteria or docs, only a running app, recording, screenshot or verbal brief, and it needs test cases. Explores the feature and applies HTSM/SFDPOT, Whittaker attacks, FEW HICCUPPS oracles and ISO 25010 to produce a test case matrix.

Use when a JavaScript or TypeScript suite has high coverage but bugs still slip through, and the team wants mutation testing to find weak tests. Sets up StrykerJS: the Jest, Vitest, Mocha or Karma runner, stryker.conf.json, thresholds and incremental runs on pull requests.

Use when a retrained ML model is a candidate for promotion and CI must block it if it got worse. Compares held-out accuracy, F1, AUC or RMSE with the production model within a tolerance, and checks key segments with Deepchecks so a global gain cannot hide a slice regression.

Use when a spec's outcome depends on several interacting conditions (pricing, eligibility, discounts, routing rules) and needs manual test cases. Builds a decision table, collapses irrelevant and impossible combinations and writes one test case per remaining column, per ISTQB.

Use when a service needs load or performance testing and no tool, load profile or pass/fail threshold is chosen yet, or a live performance incident needs its cause found. Picks k6, JMeter, Gatling, Locust or Artillery, explains the six load profiles and percentiles, and gives a first CI gate.

Use when an LLM app needs production tracing, monitoring or debugging of model calls, or production traces should feed offline eval datasets. Adds Langfuse with @observe or startActiveObservation, records metadata, cost and latency, scores traces and manages prompts as code.

Use when a C project wants Unity, CMock and CException run by one build command, or CMock mocks need writing or reading: Ceedling project scaffolding, project.yml, test tasks, JUnit XML, gcov coverage, host vs cross builds and CI. The Unity assertion API is unity-test-framework-c.

Use whenever a schema is managed with Liquibase: changelogs and changesets in XML, YAML, JSON or SQL, update, status, rollback and tag, preconditions, contexts and labels, and the changelog lock tables, for portable migrations across databases with fine rollback control.

Use when a team wants declarative, Terraform-style database schema management or migration linting beyond Flyway and Liquibase: Atlas schema apply, migrate diff against a dev database, migrate lint for destructive or locking changes, and migrate hash, across a dozen databases.

Use when a bug only appears under load on shared in-process state (cache, counter, pool, lazy singleton), or a race needs a regression test before the fix. Forces interleavings with barriers and latches, and runs ThreadSanitizer, go test -race with goleak, or jcstress for the JVM.