Use for any question or task about turning a user story, ticket or spec into test cases, including one-line stories whose gaps QA finds only after coding. Produces a markdown test case matrix (id, title, precondition, steps, expected, tier) covering happy, alternate, boundary and negative paths, agreed before automation and ready for TestRail, Qase or Xray.

Use when CI is slow or dominated by end-to-end tests, flaky failures are rising, or for a quarterly suite review. Ranks every E2E test by ROI (regressions caught vs runtime, flake rate and maintenance) and recommends which to retire, move to a lower layer or fix.

Use when a page feels slow to respond while LCP and CLS are green, or key interactions (form submit, modal open, route change) need an INP budget in CI. Breaks Interaction to Next Paint into input delay, processing and presentation, asserts per-interaction budgets in Playwright and finds long tasks.

Use whenever an Android app needs UI tests in Google's Espresso: onView with perform and check, automatic synchronization instead of sleeps, IdlingResource for app async work, ./gradlew connectedAndroidTest and JUnit XML results.

Use for any question or task about data drift or quality monitoring for a production ML model, including a fired drift alert and metrics such as precision that degrade slowly unnoticed: Evidently reports with tests for data drift, target drift and performance regression as a CI gate or scheduled job, and a playbook to pick rollback, retrain, quarantine or alert re-tuning.

Use when a feature can't or shouldn't be fully automated and a human tester needs a script: UAT sign-off, regression baselines, certification, deploy smoke checklists. Writes self-contained manual test cases as a step table, Gherkin, a business-language UAT script or a one-line checklist.

Use when preparing an app for translation, checking localization infrastructure, or when pseudo-localization comes up. Replaces strings with accented, 35%-longer variants to expose hardcoded text, truncation, encoding and bidi problems without real translators.

Use when a product ships in several languages and untranslated text keeps slipping in. Scans code for user-facing strings not wrapped in the translation function, maps gaps to the translation files, reports coverage per language and fails PRs that add new untranslated strings.

Use for any question or task about a slow SQL query or a database bottleneck, including how to read EXPLAIN or EXPLAIN ANALYZE output and database CPU that saturates under load. Reads plans from PostgreSQL, MySQL or SQLite, finds the main cost (sequential scans, nested loops, sort spills, missing or unusable indexes) and proposes the index or query rewrite.

Use when a server-streaming, client-streaming or bidirectional gRPC call needs tests, or a suite needs auditing for gaps: ordering, completion, cancellation, deadlines and partial failures, with skeletons for Go, Python, JVM and Node, plus the 17 gRPC status codes and their retry rules.

Use when a Hasura GraphQL Engine needs tests, especially role-based row permissions: a docker-compose test instance, metadata API setup, x-hasura-role and x-hasura-user-id headers, and a role by table by operation permission matrix against the v1/graphql endpoint.

Use whenever an Apollo Server v4+ GraphQL service needs tests: in-process executeOperation calls, supertest against an ephemeral port, context injection, response and error assertions, and checks that production config disables introspection and enforces persisted queries.

Use for any task that writes or runs GoogleTest tests for embedded C++ (C++17 or later) on ARM, even one test with a mock, including firmware tests that pass on the host but fail on the target: TEST, TEST_F and parameterized tests, EXPECT vs ASSERT, fixtures, GoogleMock, cross-compiling with arm-none-eabi-g++, running on host or QEMU and XML output for CI. Pure C uses unity-test-framework-c.

Use for any question about whether a GraphQL schema change is safe, including one pasted diff and a field change that broke a client: diffs schemas with graphql-inspector, classifies changes as breaking, dangerous or safe (removed fields, narrowed types, deprecated-field removal with usage checks), checks Federation supergraph drift and gates CI.

Use when a service that takes two or more locks hangs in production with no error, or before a release when lock order across code paths was never checked. Records lock order, finds cycles, uses lock timeouts to fail instead of hang, and reads jstack or gdb thread dumps.

Use when a product processing EU or California personal data needs GDPR or CCPA/CPRA readiness tests: data subject access, erasure, portability and objection, consent and withdrawal, data residency, 72-hour breach notice, data minimization, and CCPA rights including Global Privacy Control.

Use when a terminal UI (TUI) app needs tests that catch layout regressions. Captures rendered terminal frames as SVG or text snapshots and diffs them on each run: pytest-textual-snapshot for Textual, insta for Ratatui, teatest for Bubble Tea, ink-testing-library for Ink.

Use for any question or task about running tests in Jenkins, including quick ones such as running tests in parallel or publishing JUnit results when tests fail: declarative Jenkinsfile test stages, parallel stages across agents, post always/failure actions, JUnit report publishing and lockable resources for shared test infrastructure.

Use when CDN or browser caching needs tests: purges and invalidations on Cloudflare, Fastly (surrogate keys) and CloudFront, end-to-end write, purge and fresh-read checks, propagation delay and cache-status headers, plus browser Cache-Control, ETag and service-worker caching with Playwright.

Use for any question or task about bug reports, including what makes a good one, turning an informal complaint or a CI failure (JUnit XML, Allure, pytest or Playwright output) into a report, and reports closed as cannot reproduce. Fills summary, environment, steps, expected vs actual and severity rationale, then checks each field is usable.