Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Testing & quality
Comprehensive adversarial audit of a theory, proof, math/econ paper, codebase, or set of claims — decompose into components, fan out independent skeptics that must return CONCRETE defects, adjudicate every finding with a separate judge, fix all confirmed defects, then re-verify. Use when correctness must be bulletproof and single-pass or round-by-round review is too slow and too shallow. Invoke for "audit this rigorously", "find ALL the bugs/gaps", "make this rock solid", "converge faster on correctness".
Critique UI/UX designs for clarity, hierarchy, interaction, accessibility, and craft. Use for design reviews, PR feedback on UI changes, evaluating mockups, checking if a component is ship-ready, or when honest feedback is needed on whether something meets a high bar.
Run Vite Doctor diagnostics for Vite, Vue, Nitro, and Nuxt projects and use stable diagnostic codes to guide fixes. Use before review on framework projects, when Doctor/Vite Doctor/Nuxt Doctor/diagnostics are requested, or when framework-specific runtime issues are suspected.
[Beta] CI-only eval regression runner using gh-aw (GitHub Agentic Workflows). Runs all eval cases in .evals/ on a schedule or per-PR, reports pass/fail results, and can block merges on regressions. Also creates new eval cases from promoted patterns flagged by learning-aggregator-ci. Use when: you want automated regression testing of promoted rules in CI/headless pipelines. For interactive eval creation and runs, use eval-creator.
Analyze an existing website before improve, redesign, rebuild, or migration decisions. Use when a project has a current URL, local HTML/codebase, screenshots, sitemap, robots, analytics export, asset folder, or legacy site materials and needs a local-first audit with approval-gated crawling or authenticated access.
Diagnose intermittent VS Code Electron smoke-test failures from the Azure DevOps Flaky Smoke Tests pipeline (definition 700). Covers finding failed iterations, downloading task logs and platform artifacts with Azure CLI, correlating cumulative runner logs, tracing the introducing commit, and queueing focused validation runs.
Verify rendered Flutter interfaces through screenshots, golden tests, breakpoint comparisons, state matrices, and structured visual critique. Use when asked to compare designs, inspect polish, validate responsive or theme behavior, diagnose visual regressions, or confirm implementation quality after UI changes.
Use this skill when the user needs a comprehensive review of frontend code, including React, Vue, Angular or plain JavaScript/TypeScript projects. Applies to code review, Pull Request review, performance audits, security checks and similar scenarios.
Pair-programming partner skill that automatically reviews code and suggests improvements after Claude generates it. When Claude writes, modifies, or generates code for the user, this skill triggers a self-review process that checks code quality, security, performance, and best practices, and proactively points out areas for improvement. Applies to any code-generation scenario, including new feature development, bug fixes, code refactoring, API implementation, and more.
A guided interview to challenge and validate QA automation plans, test strategies, and framework designs before implementation. Use when the user wants to validate a test architecture, challenge a testing decision, prepare an AI-assisted testing rollout, or uses any "grill" trigger phrase (grill my test plan, audit my QA automation framework, challenge my QA strategy, review my test architecture).
Diagnose and fix failing GitHub Checks on PR work by default: fetch evidence, write ephemeral ci-fails report, apply fixes, verify locally. Run automatically — do not wait for user to ask. Never merge without authorize merge. Companion to /heyeddi fails.
JaCoCo Java code coverage tool USE WHEN: user mentions "JaCoCo", "Java coverage", "code coverage", asks about "coverage threshold", "jacoco-maven-plugin", "coverage report", "LINE coverage", "BRANCH coverage" DO NOT USE FOR: JavaScript/TypeScript coverage - use Vitest skill, SonarQube analysis - use `sonarqube` skill, test execution - use testing skills
JUnit 5 testing framework with Mockito integration. Covers unit tests, mocking, and Spring Boot test integration. USE WHEN: user mentions "junit", "java test", "mockito", asks about "@Test", "@Mock", "@InjectMocks", "unit test java", "spring boot test" DO NOT USE FOR: Integration tests with containers - use `testcontainers`; REST API tests - use `rest-assured`; E2E tests - use Selenium; JavaScript/TypeScript - use `vitest` or `jest`
Run this repository's local CI parity commands and `pnpm run ci:local` jobs. Use when validating a change against the same job shapes used in GitHub Actions, choosing the smallest local CI job for a package, workflow, E2E, Metro, devtools, bundle-size, or build-and-test change.
Enforce low Cognitive Complexity (SonarSource) and low Cyclomatic Complexity in ALL code written or modified, in any programming language, framework, or platform. This skill MUST activate automatically whenever code is being written, generated, modified, or refactored — no explicit trigger needed. Triggers include writing any function, method, class, module, script, handler, endpoint, test, or code block. Also triggers on "low complexity", "cognitive complexity", "cyclomatic complexity", "reduce complexity", "simplify code", "too complex", "refactor for readability", "clean code", "implement", "fix bug", "add feature", "generate test", "optimize", "rewrite", "scaffold".
Migrate configuration from Bluejay voice AI testing platform to Coval. Use when customer says "migrate from bluejay", "bluejay migration", "import bluejay config", or needs to transfer agents, simulations, metrics, and schedules from Bluejay to Coval.
Port a Selenium WebDriver or Selenide (Java) test suite to CodeceptJS 4. Trigger when the project contains `pom.xml` declaring `selenium-java` / `selenide` / `webdrivermanager` / `junit-jupiter` / `testng`, a `build.gradle{,.kts}` with `org.seleniumhq.selenium` / `com.codeborne:selenide`, a `src/test/java/` tree with `*Test.java` / `*IT.java` / `*Steps.java`, imports from `org.openqa.selenium.*` (`WebDriver`, `By`, `WebDriverWait`, `ExpectedConditions`, `Actions`) or `com.codeborne.selenide.*` (`Selenide`, `SelenideElement`, `Condition`), JUnit/TestNG annotations (`@Test`, `@BeforeEach`, `@DataProvider`, `@FindBy`, `@ParameterizedTest`), `PageFactory.initElements(...)`, `new ChromeDriver(...)` / `Selenide.open(...)`, or Cucumber-JVM step defs (`@Given` / `@When` / `@Then`).
Generates multi-user workflow documentation by interviewing the user about personas, exploring the codebase for multi-user patterns, then walking through the live app with per-persona Playwright CLI named sessions to co-author interleaved, persona-tagged workflows. Use when the user says "generate multi-user workflows", "create multi-user workflows", or "generate concurrent user workflows". Produces persona-tagged workflow markdown that feeds into the multi-user converter and Playwright runner.
Run and interpret the OnionCry Rust CLI for architecture-boundary checks in JavaScript and TypeScript repositories. Use when the user asks to check architecture boundaries, generate or review .onioncryrc.jsonc, run onioncry check or explain, adopt a violation baseline, inspect llm-mode reports, render boundary graphs, add repo/package scripts, or automate OnionCry in agent workflows. Do NOT use for generic Rust implementation work.
Analyze web and backend performance using Web Vitals, network timing, and Node.js metrics. Use when the user asks about page performance, load times, Core Web Vitals (LCP, CLS, INP), slow pages, backend bottlenecks, or SEO performance factors.
Complete browser automation with Playwright. Auto-detects dev servers, writes clean test scripts to /tmp. Test pages, fill forms, take screenshots, check responsive design, validate UX, test login flows, check links, automate any browser task. Use when user wants to test websites, automate browser interactions, validate web functionality, or perform any browser-based testing.
Use when managing prompt templates, running evaluations, building workflows, logging LLM requests, or analyzing AI application performance. Reach for this skill when working with prompt versioning, release management, observability, tables/datasets, or multi-step AI workflows.
Find, rank, and safely remove accidental codebase complexity by proving real consumers, dynamic entrypoints, compatibility obligations, duplicate representations, speculative surfaces, and lifecycle ownership. Use when asked to simplify or clean up a repository, reclaim code entropy, reduce over-engineering or redundancy, find deletion candidates, collapse duplicate state/APIs, remove dead or added-then-abandoned code beyond static-tool output, or implement an evidence-backed simplification pass in any language or stack. Also trigger for 代码化简、熵回收、删代码、清理冗余、收敛抽象、去除过度设计. Do not use as a performance audit unless simplification is the stated goal.
Comprehensive skill for 89 refactoring techniques and 22 code smells with practical examples. Use when the user asks to refactor code, detect code smells, improve code quality, reduce complexity, or clean up technical debt. Covers composing methods, moving features between objects, organizing data, simplifying conditionals and method calls, dealing with generalization, and detecting smells across bloaters, OO abusers, change preventers, dispensables, and couplers with before/after comparisons and step-by-step mechanics.
Perform a security review of a diff, branch, or pull request, assessing what the change introduces, weakens, or exposes, with a triage-first workflow and false-positive discipline. Use when asked to security review a PR or branch, check a diff before merge or commit, or assess whether a change is safe to ship
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.