Use for any question or task about testing Stripe payments, including one webhook handler and checkout that fails for 3D Secure cards: test-mode keys, test cards for success, declines and 3D Secure, stripe listen to forward webhooks locally, stripe trigger fixtures, and Stripe-Signature verification.

Use for any question or task about testing push notifications, including iOS and Android pushes and one FCM or APNs sender: Firebase FCM, Apple APNs and Web Push with VAPID, payload encryption, badge, sound and click actions, expired or unregistered tokens, silent vs alert pushes and topic routing, plus in-app delivery over WebSocket or SSE with read state.

Use for any question about mutation testing in C or C++, or when a C/C++ suite has high coverage but bugs still ship: sets up Mull for Clang builds (version-matched package, -fpass-plugin flags, CMake, mull-runner, mutators, path filters) and runs it in GitHub Actions. Other languages: stryker-mutation, stryker-net-mutation, pitest-mutation, mutmut-mutation.

Use when a tenant onboarding or offboarding flow is added or changed, or before a compliance review: tests for account creation, isolation from the first API call, default quotas, billing links, seed data, idempotent re-provisioning and full deletion on teardown.

Use whenever a React Native app needs end-to-end tests with Detox: gray-box runs inside the app, by.id, by.text and by.label matchers, waitFor for extra sync, and the Jest runner. Flutter is flutter-testing; cross-platform black-box is appium-testing; YAML flows are maestro-flows.

Use when a project already uses Apache JMeter or needs JVM-native load testing: .jmx test plans with thread groups, HTTP samplers, assertions and listeners, headless jmeter -n runs, the HTML dashboard and CI gates from JTL results.

Use for any question or task about RAG or agent evaluation metrics, including which metrics to use, scoring an answer against its contexts, and telling retrieval problems from generation problems after a change such as chunk size: Ragas faithfulness, response relevancy, context precision and recall, noise sensitivity, factual correctness, tool-call accuracy, agent goal accuracy and rubric scoring.

Use when one test is known to fail intermittently and reading its code has not explained why. Varies one condition at a time (isolation, order, workers, viewport, latency, repetition), counts passes and failures with confidence intervals, and walks shared resources once parallelism is implicated.

Use when an OpenTelemetry Collector config changes (a new receiver, processor or exporter) and must be verified before deploy. Runs otelcol validate, uses the debug or file exporter to assert spans flow through the pipeline, and wires the check into CI.

Use when a native Windows app (UWP, WPF, WinForms, Win32) must be driven from a Selenium-style client in C#, Java, Python, Ruby or JS: WinAppDriver directly or through the Appium 2 Windows driver, capabilities, AccessibilityId and Name locators, and Windows CI runners. C#-only: flaui-tests.

Use for any question about how to test or automate a desktop GUI app on Windows, macOS or Linux, including which tool to pick for WPF, WinForms, Qt or Electron apps and tests that break because they click by coordinates. Maps each app to a driver (FlaUI, WinAppDriver, Playwright for Electron, QtTest, XCUITest, AT-SPI) and covers accessibility-first locators, waits and elevation hazards.

Use for any question about whether an OpenAPI or Swagger spec change breaks clients, including one pasted diff and a partner integration that broke: removed endpoints, deleted schemas, narrowed enums, changed responses. Runs oasdiff breaking, grades findings ERR, WARN or INFO and gates CI with --fail-on.

Use when a team wants maintainability and technical-debt ratings to gate PRs, or a SonarQube project reports numbers nobody tied to a gate. Runs SonarQube or SonarCloud on production code for code smells, bugs and maintainability ratings, the maintainability view rather than security.

Use when a Chrome extension's messages reach no listener, sendResponse never fires, a port disconnects mid-test, or a web page's access must be proven. Tests runtime.sendMessage, tabs.sendMessage, long-lived ports, externally_connectable and native messaging on a running extension.

Use when cron jobs, Kubernetes CronJobs, BullMQ repeat jobs, Sidekiq schedules or any time-based job need tests: cron expression validation, DST and leap-day edge cases, catch-up after downtime, overlapping-run locks with stale-lock recovery, and time zones.

Use when a complex stateful flow (checkout, onboarding, multi-step wizard) needs systematic model-based coverage that ad-hoc tests miss. Builds a state-machine model, checks it is complete, picks a coverage criterion, generates covering paths and turns them into a curated test suite.

Use after an accessibility scan across many pages (pa11y-ci, axe over a sitemap, Lighthouse batch) when the team needs a shareable conformance report. Groups violations by WCAG 2.2 success criterion and level, flags pages that were not scanned and writes a summary plus per-page detail.

Use for any question or task about Cypress tests, including quick how-to questions and one pasted test to fix: stubbing API calls with cy.intercept, replacing cy.wait sleeps with assertions, tests that pass locally but fail in CI, fixtures and cy.task, cy.clock, cy.origin and cy.session logins, cypress.config.ts, the Cypress 16 upgrade and Cloud parallel runs.

Validate and write a concise constraint for a project or product. Use when Codex needs to turn an external platform limitation, technical invariant, policy, dependency, or other fixed condition into a precise constraint with concrete implications and authoritative sources.

Guide and audit Node.js CLI application development against 41 established best practices covering UX, distribution, interoperability, accessibility, testing, error handling, development setup, analytics, versioning, and security. Use this skill when building, extending, reviewing, or scaffolding a Node.js CLI — including when someone says "audit my CLI", "review my CLI code", "I'm building a CLI tool", or asks about adding argument parsing, help output, shell completion/autocomplete, error handling, color output, STDIN, --json flags, exit codes, --version flags, or npm publishing. Applies even when best practices are not explicitly mentioned. Also trigger for "how should I implement X in my CLI" or "what's the right way to do Y in a Node.js CLI". Do NOT use for Node.js backend or API development with no CLI entry point.