Migrates Cypress, Selenium, WebdriverIO, or Protractor suites to Playwright in staged slices, each with a parity gate before the old test is deleted. Use when porting a legacy browser suite, when deciding which tests are not worth porting, when a half-finished migration has stalled with two suites running in parallel, or when a suite migration needs a plan before anyone starts translating files.
日本語の概要は準備中です。原文の説明を表示しています。
Decides what to fake and stubs it correctly: network interception with Playwright route or HAR replay, MSW handlers, fake clocks, and fixed timezones. Use when a test depends on a third party, when a date-sensitive test breaks overnight, when a suite is slow because it calls real services, or when a mocked test stays green while production is broken.
日本語の概要は準備中です。原文の説明を表示しています。
Runs session-based exploratory testing: writes charters, timeboxes sessions, applies coverage heuristics and tours, captures notes as evidence, debriefs, and converts findings into bug reports and automation candidates. Use when a feature needs testing before requirements settle, when scripted cases keep passing while users hit problems, when a release needs a risk sweep with limited time, or when the request mentions charters or exploratory sessions.
日本語の概要は準備中です。原文の説明を表示しています。
Produces production-ready Product Requirements Documents (PRDs) for software systems and AI-powered features, with clear problem framing, measurable outcomes, scoped functionality, testable requirements, and explicit risks. Use when the user wants to write a PRD, define requirements, plan a feature, or turn a vague product idea into an implementation-ready specification.
日本語の概要は準備中です。原文の説明を表示しています。
Requires an externalised rationale for high-risk changes - new public exports, new endpoints, auth edits, migrations, removed guards - recorded as an Intent commit trailer, an ADR reference, or a module intent register, and reports high-risk changes that carry none. Use when agent-generated or AI-assisted changes ship without a recorded why, when reviewers cannot tell what a diff was for, when a codebase is losing its decision history, or when setting up an intent gate alongside test and coverage gates.
日本語の概要は準備中です。原文の説明を表示しています。
Transforms rough tester notes, screenshots, console output, or observed behavior into reproducible defect reports with severity, evidence, and follow-up guidance. Use when logging bugs, triaging intermittent issues, or rewriting vague defect notes into developer-ready reports.
日本語の概要は準備中です。原文の説明を表示しています。
Maps requirements (PRD, user stories, acceptance criteria) to planned test coverage via a Requirements Traceability Matrix, exposing coverage gaps, risks, test levels, prioritization, and automation candidates. Use when designing coverage from a specification, checking coverage completeness for a PRD or user story, finding missing acceptance criteria, or building a risk-based regression strategy. When the tests already exist and the matrix has to be extracted from them, verified, and kept accurate, use the tracing-requirements-to-code skill instead.
日本語の概要は準備中です。原文の説明を表示しています。
Human review protocol for whether AI output is grounded in its sources: claim-by-claim attribution, hallucination classes, citation verification, sampling that is defensible, and a recorded attestation because groundedness cannot be fully automated. Also covers the responsible-AI pass - harmful output, bias, disclosure and privacy. Use when reviewing a RAG or summarisation feature, when factuality matters more than fluency, or when an eval suite needs the human check it cannot replace.
日本語の概要は準備中です。原文の説明を表示しています。
Sets up and maintains visual regression testing: what to snapshot, baseline strategy, masking dynamic regions, threshold tuning, containerized baselines, and the review-and-update workflow. Use when styling regressions escape to production, when snapshots fail on every machine or every run, when baselines are being updated without being looked at, or when deciding whether visual testing is the right tool at all.
日本語の概要は準備中です。原文の説明を表示しています。
Classifies a diff into file tags and hunk tags - new public export, new endpoint, modified auth, SQL string, migration, added dependency, touched prompt - then maps each tag to the quality practices it makes relevant, producing a defensible per-change check scope instead of running everything or guessing. Use when deciding what to test for a specific pull request, when a full regression run is too slow to gate on, when a pre-push or PR gate needs a scope someone can argue with, or when asked "which checks does this change actually need".
日本語の概要は準備中です。原文の説明を表示しています。
Creates, reviews, and modernizes static code analysis setups for Node.js and TypeScript repositories, covering ESLint flat config, typescript-eslint, tsconfig, Prettier, import sorting, Husky, lint-staged, package.json quality scripts, and CI quality gates. Use when setting up or auditing linting, formatting, type-checking, commit hooks, or GitHub Actions quality checks in a TypeScript project.
日本語の概要は準備中です。原文の説明を表示しています。
Analyzes technical debt in codebases, test suites, architecture, dependencies, and delivery workflows using observable signals. Use when auditing repository health, explaining slow delivery or flaky tests, prioritizing refactoring, or building an evidence-based remediation roadmap with risk, effort, and ROI. Use when user asks for technical debt analysis, repository audit, or refactor planning.
日本語の概要は準備中です。原文の説明を表示しています。
Runs the red-green-refactor loop with a compact set of rules: one failing test, the minimal code to pass it, cleanup under green. Use for routine test-first work on a single behavior or a small bug fix, and when the request mentions TDD, test-first, red-green-refactor, or writing the test before the code. For multi-unit features, outside-in double loops, legacy code without tests, or a session that needs a written cycle log, use the test-driven-development skill instead.
日本語の概要は準備中です。原文の説明を表示しています。
Drives implementation test-first through red-green-refactor cycles: one failing test, the minimal code to pass it, then cleanup under green. Use before writing production code for a new behavior, and whenever the request mentions TDD, test-driven, test-first, red-green-refactor, "write the test first", "start with a failing test", or reproducing a bug with a test before fixing it. Also use when implementation keeps landing before anyone knows how it will be verified.
日本語の概要は準備中です。原文の説明を表示しています。
Tests the customization assets themselves - skills, prompts, custom agents, instructions - the way a product is tested: activation cases that check an asset fires when it should and stays quiet when it should not, output-contract cases, safety cases, collision cases between assets competing for the same trigger, a weighted rubric scored blind, and a baseline-versus-candidate gate before an edit ships. Use when a skill is edited and nobody knows whether behaviour changed, when two skills fight over the same request, when a description is being tuned for discoverability, when a collection has grown past manual spot-checking, or when the request mentions skill evals, prompt regression, or "does this skill actually work".
日本語の概要は準備中です。原文の説明を表示しています。
Validates API responses against OpenAPI or JSON Schema, detects breaking changes between spec versions, and builds consumer-driven contract checks. Use when an API has a published spec, when a backend change might break a client, when API tests assert only status codes, or when mocked fixtures need a guard against drifting from the real service.
日本語の概要は準備中です。原文の説明を表示しています。
Applies OWASP-informed security testing within a tester scope: an authorization gate before any probing, authorization matrices, broken access control and IDOR checks, input validation and injection observation, session and auth behaviour, and safe reporting of findings. Use when testing auth-protected features, when a security review needs QA coverage, when an exploratory session surfaced something security-shaped, or when access control needs a systematic check rather than a spot check.
日本語の概要は準備中です。原文の説明を表示しています。
Builds an offline eval suite for LLM-powered features: golden reference cases per prompt template, deterministic scoring where possible, a regression gate on prompt, model and retrieval changes, and CI wiring with promptfoo, Vitest or deepeval. Use when a product ships prompts, agents or RAG, when a model or prompt upgrade needs a regression check, when LLM output is currently verified by someone eyeballing it, or when asked how to test a feature whose output is non-deterministic.
日本語の概要は準備中です。原文の説明を表示しています。
Tests the runtime validation around model output - schema conformance, PII leakage, moderation, refusal behaviour, tool-call safety - and the adversarial suite that tries to defeat it: direct prompt injection, jailbreaks, and indirect injection through retrieved documents, tool results and user content. Use when an LLM feature reaches real users, when a model can call tools or read untrusted content, when an AI feature needs a security-style test suite, or when asked whether a chatbot or agent can be manipulated.
日本語の概要は準備中です。原文の説明を表示しています。
Designs and runs performance and load tests: workload modelling from real traffic, thresholds tied to SLOs, warmup and ramp shapes, percentile-based analysis, and lightweight CI perf checks with k6 or Artillery. Use when a feature has latency or throughput requirements, when "it feels slow" needs to become a number, when a launch needs a capacity check, or when a performance result needs interpreting rather than just collecting.
日本語の概要は準備中です。原文の説明を表示しています。
Builds and maintains bidirectional traceability between requirements and the tests that verify them: extracts a matrix from an existing codebase, annotates tests with requirement IDs, finds orphan tests and uncovered requirements, verifies that each link is real, and enforces linkage in CI. Use when a suite exists but nobody can say what it proves, when an auditor or stakeholder asks which tests cover a requirement, when a traceability matrix has gone stale, or when a requirement changes and its blast radius must be found.
日本語の概要は準備中です。原文の説明を表示しています。
Defines, computes, and interprets test and quality metrics: pass rate, flake rate, suite duration, defect escape rate, time to detect, and coverage with its caveats. Use when building a QA dashboard, reporting suite health to stakeholders, comparing releases over time, or when a coverage percentage or a bug count is being treated as a measure of quality.
日本語の概要は準備中です。原文の説明を表示しています。
Cuts AI tells from what the agent says about its own work: completion claimed without a run, invented file paths and API names, terminal output that was never produced, findings with no location, percentages with no denominator, absence claimed from one grep, effort narration instead of results, hedge stacking, buried blockers, puffery in a bug title. Use whenever reporting a result, a fix, a review finding, a test outcome, a coverage or flake number, a root cause, or a release recommendation, and when the request mentions "prove it", "did you actually run it", "be specific", "no fluff", "stop hedging", or "is this real". Must always apply to answers about your own work.
日本語の概要は準備中です。原文の説明を表示しています。
Cuts AI tells from test code: tests that pass without proving anything, tautological assertions, mock-only tests, hardcoded waits, coverage theater, vague names, swallowed errors, retries used as fixes. Use whenever test code is written, changed, or reviewed, including tests produced as a side effect of a feature task, and when the request mentions "review these tests", "are these tests any good", "this test always passes", "this suite is flaky", or "clean up these tests". Must always apply to test code.
日本語の概要は準備中です。原文の説明を表示しています。