| Accessibility audit - deep dive |
Run a deeper accessibility audit for a page, sampled set of pages, or user flow with WCAG 2.2 mapping, manual verification guidance, and technical or stakeholder-ready output. For a quick single-URL check, use the a11y-webpage-audit prompt instead. |
  |
| A11y webpage audit (single URL) |
Analyze one webpage for accessibility with WCAG 2.1/2.2 mapping and actionable fixes. Quick single-URL audit; use the a11y-audit-deep-dive prompt for multi-page flows and stakeholder-ready reporting. |
|
| Analyze test results |
Turn raw test run output - a CI log, JUnit/Playwright report, or console output - into an actionable summary: failure clusters, likely root-cause groups, flakiness suspects, and what to investigate first. |
|
| API test plan and test generator |
Create a risk-based API test plan and generate example automated tests from API definitions (OpenAPI/Postman/custom docs). API-focused end-to-end flow: plan first, then tests. |
|
| Audit collection health |
Audits a customization collection as a whole: inventory, frontmatter and structural conformance against the house conventions, description trigger quality, overlap and collision clusters between near neighbours, broken cross-references, orphaned and missing resources, drift between assets, plugin copies and generated docs, and which assets have no behavioural test. |
|
| Audit quality debt and silent skips |
Audits the three debts - technical, intent and comprehension - and inventories every silent skip in the repo: skipped tests, disabled lint rules, ts-expect-error, continue-on-error, coverage ignores, lowered thresholds, raised retries and expired waivers. |
|
| Create a bug report |
Turn rough tester notes, screenshots, logs, or observed behavior into a professional bug report with severity, reproducibility, and evidence guidance. |
|
| Create a new skill |
Create a new skill (SKILL.md) for VS Code Copilot. Analyzes existing skills for patterns, guides through design decisions, creates the skill file with supporting resources, and validates against best practices. |
|
| Derive a quality contract |
Derives which testing practices are MUST, SHOULD or COULD for this specific project from its risk profile, team maturity and product surfaces, then reports a PRESENT/PARTIAL/MISSING/WAIVED gap matrix and an ordered remediation plan. |
|
| Derive test cases from code |
Read a function, module, or diff and derive white-box test cases from its actual branches, boundaries, and error paths - then compare against existing tests to expose coverage gaps. |
|
| Design an LLM eval suite |
Designs and implements an offline eval suite for an LLM feature: golden, edge and negative cases per prompt template, the cheapest scoring strategy that can actually fail, pinned determinism, a regression gate against a committed baseline, and the CI job that runs it. |
|
| Analyze and explain code |
Analyze and explain the selected code, the active file, or a provided snippet: purpose, structure, patterns, risks, and what a tester should verify |
|
| Explain a test failure |
Diagnose a failing test from an error message, stack trace, CI log, or trace file: classify the root cause as app defect, test bug, environment issue, or flakiness - without changing any code. Use the fix-tests prompt when you want the failure fixed. |
|
| Fix failing tests |
Diagnose and fix failing tests by addressing root causes, without weakening assertions or masking defects |
|
| Generate exploratory testing charters |
Create prioritized, timeboxed exploratory testing session charters from a feature description, release notes, or live app exploration - with risk focus, test ideas, and oracle notes for each session. |
|
| Generate Gherkin scenarios |
Convert a user story, acceptance criteria, or feature description into well-formed Gherkin scenarios (Given/When/Then) with scenario outlines for data variations, ready for BDD frameworks like Cucumber or Playwright-BDD. |
|
| Generate test data |
Design realistic and edge-covering test data packs from requirements, constraints, or existing test plans for manual and automated testing. |
|
| Generate manual test cases |
Turn a feature description, acceptance criteria, exploratory notes, or a test plan into detailed manual test cases with risk tags, expected results, and automation hints. |
|
| Map requirements to test coverage |
Build a Requirements Traceability Matrix from a PRD, user stories, or acceptance criteria against existing tests: what is covered, at which level, what is missing, and which gaps carry the most risk. |
|
| Migrate tests to Playwright |
Migrate existing UI tests from Cypress, Selenium, Puppeteer, or Protractor to Playwright while preserving coverage: translate idioms to Playwright best practices, run both suites where possible, and report an honest migration status per test. |
|
| Performance and reliability test planner |
Design a performance and reliability test strategy and propose concrete load/soak tests for critical user flows. |
|
| Explore a website and gather network requests using Playwright MCP |
Explore a website with Playwright MCP while capturing network requests and responses. Use the playwright-explore-website prompt instead when network traffic is not needed. |
|
| Explore a website and propose test cases |
Explore a website with Playwright MCP and propose test cases. Use the playwright-explore-website-requests prompt instead when network traffic should also be captured. |
|
| Generate tests based on a scenario using Playwright MCP |
Generate a Playwright test from a scenario description by first executing it live via Playwright MCP. Use the test-generator prompt instead when generating from a written test plan file. |
|
| QA strategy - edge cases, security and attack scenarios |
For any feature, user story, or API spec: instantly generate a structured scenario matrix covering edge cases, boundary values, OWASP Top 10 security attacks, and adversarial sequences. |
|
| Record the intent behind a change |
Finds the high-risk surface in a change range, checks whether a rationale already exists in commit trailers, ADRs or a module intent register, and drafts an Intent trailer for the author to confirm - why now, what constrained the shape, what was rejected. |
|
| Refactor tests to page objects |
Refactor UI test specs into a clean Page Object Model without changing what the tests verify: extract locators and actions, keep assertions in tests, and prove behavior is unchanged by rerunning the suite. |
|
| Analyze regression scope |
Turn a diff, PR summary, changed files, or hotfix description into a prioritized regression scope with a minimal confidence suite and retest guidance. |
|
| Review requirements for testability |
Review a user story, PRD, or acceptance criteria for ambiguity, missing states, and untestable statements - before implementation starts. Produces concrete clarifying questions and rewritten, testable acceptance criteria. |
|
| Scope and gate this change |
Tags a diff into file and hunk tags, resolves which quality practices the change makes relevant, verifies diff coverage and intent rationale, and returns severity-ranked findings with implementable remediation briefs. |
|
| Stabilize flaky tests |
Hunt down intermittently failing tests, classify the flakiness root cause (timing, environment, data, state leakage), fix it properly, and prove stability with repeated runs. Use the fix-tests prompt for tests that fail consistently. |
|
| Tech debt audit - test automation code review |
Ruthlessly audit automated test code for anti-patterns: fragile selectors, hardcoded waits, missing assertions, test interdependencies, credential leaks, and more. Produces a severity-ranked debt report with line-level citations and concrete fixes. |
|
| Generate tests based on test plan |
Generate automated tests from a written test plan file, matching the project’s existing framework. Use the playwright-generate-test prompt instead when starting from a scenario description. |
|
| Test LLM guardrails and injection resistance |
Maps the trust boundaries of an LLM feature, tests the runtime guardrails around model output - schema, PII, moderation, rendering, tool authorization, resource caps, fail-closed behaviour - and runs an adversarial corpus through every content channel including retrieval and tool results. |
|
| Generate a basic test plan |
Generate a quick, single-pass test plan for a website from light exploration. Use the test-planner prompt for a comprehensive, interactive plan covering web and API. |
|
| Generate a comprehensive test plan |
Collects environment details and produces a prioritized test plan with web and API scenarios. Use the test-plan-basic prompt for a quick single-pass plan without deep exploration. |
|
| Verify acceptance criteria |
Compare implementation evidence against acceptance criteria and show what is met, partial, missing, or still untestable. |
|
评论