Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Testing & quality
Act on a codependix run that failed or reported drift — a boundary violation, a --check reports run that found stale exports, a project that failed its export, a missing anchor block, a NestJS container that failed to boot, a rejected command line, or a --write that produced no files at all. Use when codependix exits non-zero, when a codependix-gate target fails, when an edge or a cycle breaks a declared rule, when a committed Mermaid block disagrees with a fresh run, when a graph export is missing for one project but not others, or before hand-editing an exported diagram or loosening a rule to make a check pass.
Write or edit a codometer.config.ts, declaring inputs, limits, custom convention counters, comment budgets, exclusions, or output destinations. Use when a repository has no codometer configuration yet, when adding a size or count limit, when bounding how long a comment block or a JSDoc comment may run, when declaring an input for compiled or generated output, when counting a naming convention no built-in analyzer knows about, when a limit fails to bind or an input matches no files, or when deciding whether one configuration can describe every project in a workspace
Run a codometer measurement, decide where its report goes, or read a badge block or JSON report it produced. Use when running codometer or npx @codometer/cli, choosing --output-json/--output-markdown and --check flags, wiring codometer into a CI step, splicing a badge block into a README, or reading a codometer JSON report's metrics, limits, or failures. Covers the command-line host directly, without assuming any task runner
Act on a codometer limit breach, a comment budget breach, a stale committed report, or a run that failed outright. Use when codometer --check exits non-zero, when a size or count limit is breached, when a comment block runs past its word, line, or character budget, when a committed report reads stale against a fresh measurement even though nothing changed, when a limit fails to bind or an input matches no files, or before reaching for a repository's limit value to make a failing check pass
Analyzes coupling between modules using the three-dimensional model (strength, distance, volatility) from "Balancing Coupling in Software Design". Use when asking "are these modules too coupled?", "show me dependencies", "analyze integration quality", "which modules should I decouple?", "coupling report", or evaluating architectural health. Do NOT use for domain boundary analysis (use domain-analysis) or component sizing (use component-identification-sizing).
Merges GitHub Pull Requests after validating pre-merge checklist. Use when user wants to merge PR, close PR, finalize PR, complete merge, approve and merge, or execute merge. Runs pre-merge validation (tests, lint, CI, comments), confirms with user, merges with proper format, handles post-merge cleanup.
Implement a plan, spec, or issue into working code in straight-through, tests-after, or TDD mode, then run the repo's test and build gate before declaring done. Use when the user says "implement this plan", "implement #42", "do this TDD", "build it, then add tests", "apply these review findings", or "make this failing test pass". It runs attended; the unattended issue-to-PR span is afkkit's.
Verify that a web app change actually works by driving the running app from the inside (DOM, network, routing, console, framework state) instead of screenshots or guessing. Use after any user-facing change, when a fix is claimed but unproven, when a test passes but the UI is broken, or when you need a real verdict rather than "looks right". Also use to install and wire up Reticle in a project that does not have it yet.
Generates CI/CD pipeline configurations for test automation with GitHub Actions, Jenkins, GitLab CI, and Azure DevOps. Includes TestMu AI cloud integration. Use when the user mentions "CI/CD", "pipeline", "GitHub Actions", "Jenkins", or "GitLab CI". Triggers on: "CI/CD", "pipeline", "GitHub Actions", "Jenkins", "GitLab CI", "Azure DevOps", "automated testing pipeline".
Testing practices for iOS 26 / Swift 6.2 clinic modular MVVM-C applications. Covers unit/UI/snapshot testing, protocol-based mocks, async actor isolation, and dependency-injected test architecture aligned with Domain protocols, App-target composition, and Data-owned I/O boundaries. Use when writing, reviewing, or refactoring tests for ios-* and swift-* clinic modules.
Tell callidescope what to do — the command-line flags (--check, --write, --addresses, --config, --format, and an override for every configured field, from --directories and --exclude through --include-tests, --maximum-depth, --maximum-breadth, --json, --markdown, and --mermaid) and the callidescope.config.ts they layer over, covering the two depth and breadth limits, entry points, exclusions and excluded callees, and where a run writes its JSON, markdown, and mermaid reports. Use when wiring a depth gate into CI or a commit hook, when a whole-workspace run is slow, when choosing between --check and --write, when a repository has no callidescope configuration yet, when a traced project has no callidescope.config.ts of its own or one that leaves a field out, when a project needs its own depth or breadth limit, when a package low in the graph measures nothing, when a trace judges code it should not, when everything is reported as an orphan root, or when deciding where a committed report should live.
Run callidescope and read what it printed — a whole-workspace trace, one project's gate verdict, or the depth and breadth commands against one or more callables, each addressed as file#qualified-name. Use when running callidescope or npx callidescope, when reading a call stack or a breadth row, when reading what a per-project gate or trace target printed, when reading a project readme's Limits table, when a depth printed as "≥ n" needs interpreting, when reading a committed markdown report, mermaid diagram, or JSON report, when a run narrowed with --directories reports a frame, a stack, or a project the run was never pointed at, when a depth moved without an edit that explains it, or when asking who calls this, what does it call, what would this rename touch, and where should this callable be split before a refactor starts.
Designs chaos experiments, creates failure injection frameworks, and facilitates game day exercises for distributed systems — producing runbooks, experiment manifests, rollback procedures, and post-mortem templates. Use when designing chaos experiments, implementing failure injection frameworks, or conducting game day exercises. Invoke for chaos experiments, resilience testing, blast radius control, game days, antifragile systems, fault injection, Chaos Monkey, Litmus Chaos
Generate realistic synthetic data using Spark + Faker (strongly recommended). Supports serverless execution, multiple output formats (Parquet/JSON/CSV/Delta), and scales from thousands to millions of rows. For small datasets (<10K rows), can optionally generate locally and upload to volumes. Use when user mentions 'synthetic data', 'test data', 'generate data', 'demo dataset', 'Faker', or 'sample data'.
Reviews code implemented by AI agents in four ordered passes (convention fit, agent-slop signatures, requirement completeness, correctness) against the working tree or branch diff. Findings are ranked by severity and supported by quoted evidence. Use when the user says “review this code,” “review my changes,” or “check the agent's work” before a commit or PR.
Generates a self-contained Python experiment client that uses the ddtrace.llmobs SDK. Emits either a runnable .py script or a Jupyter .ipynb notebook matching the canonical DataDog reference notebook style. Use when the user says "generate Python experiment", "write an SDK experiment", "create a ddtrace experiment", "Python notebook experiment", "use the Agent Observability SDK", or has `ddtrace` installed and wants idiomatic SDK code.
Act on a callidescope run that failed—a per-project gate reporting a stack over that project's limit, a breadth gate, a gate that read none of its project's own files, a stale committed report, an unreadable tsconfig, a run that traced nothing, or a rejected command line or configuration. Use when callidescope or an inferred gate target exits non-zero; when a stack got deeper in a change; when deciding whether a failing gate is a code fix or a limit fix; when choosing or moving a project's own maximumDepth or maximumBreadth; when a printed maximumDepth disagrees with the number a gate judged; when a committed report disagrees with a fresh run; when a declared entry-point address resolves to nothing or several targets; when a traced project has no callidescope.config.ts of its own or one that omits a field; or before reaching for maximumDepth to make a failing check pass
Senior QMS audit expert for internal and external quality management system audits. Provides ISO 13485 audit expertise, audit program management, nonconformity identification, and corrective action verification. Use for internal audit planning, external audit preparation, audit execution, and audit follow-up activities
Technical due diligence for M&A, investment, or acquisition. Reads a target company's codebase and generates a comprehensive tech DD report with architecture assessment, tech debt quantification, scalability analysis, security posture, team capability inference, build system quality, test coverage, deployment maturity, and open source license risks. Outputs tech-dd-report.md formatted like a real investment memo with risk ratings, remediation costs, and go/no-go recommendation.
Verify a released archon binary works end-to-end via a specific install path. Use when: cutting a new release, reproducing a user bug report on the released version, or validating that a hotfix binary actually works after a re-tag. Triggers: "test the release", "test 0.3.1 via brew", "verify the curl install", "smoke test the binary", "did the release binary work", "run /test-release", "verify the release". NOT for: testing dev work (use bun link directly), testing unreleased changes (build locally via scripts/build-binaries.sh first), or running the full validate suite (bun run validate is separate).
This skill should be used when debugging frontend/UI bugs that need runtime evidence. USE THIS SKILL (instead of adding console.log) when you're about to: "add console.log and ask user to check", "open DevTools and tell me what you see", "reproduce the bug and share the output", "check the browser console". Triggers: "debug this", "fix this bug", "why isn't this working", "investigate this issue", "trace the problem", "figure out why X happens", "UI not updating", "state is wrong", "value is null/undefined", "click doesn't work", "modal not showing". Automates log collection server-side - you read logs directly, no user copy-paste needed.
Eval and improve a skill against golden cases — run the target skill blind in a fresh, context-free subagent on each example input, grade the artifact against the expected outcome, and let the gaps drive the edits. Use when the user wants to test/eval/improve/harden a skill, says "this skill keeps producing X / keeps missing Y", or hands a skill plus example input→expected-output pairs. Pairs with [write-skills](../write-skills/SKILL.md) (the authoring principles every fix obeys).
Creates comprehensive GitHub Actions CI/CD workflows for linting, testing, building, and deploying. Includes caching strategies, matrix builds, artifact handling, and failure diagnostics. Use for "GitHub Actions", "CI pipeline", "workflow automation", or "continuous integration".
Audit a Gradle project against the official Gradle best practices, report findings, and propose code changes to bring the build into compliance. Use this skill whenever the user wants to check, audit, lint, review, or validate a Gradle build — and also when they want to apply, adopt, fix, or enforce best practices. Trigger on phrases like 'check my build', 'audit gradle', 'best practices review', 'any issues with my build?', 'gradle health check', 'lint my build files', 'is my gradle setup correct?', 'apply gradle best practices', 'fix my build to follow best practices', 'make my build follow best practices', 'modernize my build', or simply 'check best practices'. Also trigger when the user asks about Gradle build quality concerns such as dependency management, build performance, build structure, or task wiring.
Show a frontend change in a real browser or prove it for a PR: drive the feature, capture screenshots (plus a short GIF as proof), and publish the proof so a pull request can embed it inline; or set up the browser driver and publishing path on a machine. Use when a frontend change is built and you want to see it or get visual evidence that it works before opening the PR: “show me what it looks like,” “screenshot this,” “verify this feature,” “record the flow working,” “prove the UI change,” “/verifykit,” when a PR needs proof artifacts attached, “set up verifykit,” or “install the browser driver.”
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to “open a website”, “fill out a form”, “click a button”, “take a screenshot”, “scrape data from a page”, “test this web app”, “login to a site”, “automate browser actions”, or any task requiring programmatic web interaction.
Diagnose GitHub bug reports in openai/codex. Use when given a GitHub issue URL from openai/codex and asked to decide next steps, such as verifying against the repo, requesting more info, or explaining why it is not a bug; follow any additional user-provided instructions.
Use when creating or updating component fixtures for screenshot testing or their shared infrastructure, or when designing UI components to be fixture-friendly. Covers fixture file structure, theming, service setup, CSS scoping, async rendering, validation, and common pitfalls
Criticism and self-criticism: honestly, specifically, and factually review the work and process after completion, at a review checkpoint, when receiving criticism, or when the same mistake recurs; produce actionable improvements and address external criticism without defensiveness. Trigger signals include review, retrospective, audit, "帮我看看有没有问题", "你确定吗", "check this for problems", and "are you sure". Do not trigger at the start of a task or for a one-step lookup.
Develop a new ESLint rule for @sap-ux/eslint-plugin-fiori-tools. Always use this skill when adding any rule to the plugin, implementing manifest.json/XML/CDS/flex-change validation, writing rule tests with project context, or creating rule documentation — even if the request sounds simple or the user doesn't mention 'ESLint' explicitly. Guides through the complete workflow: diagnostics constant, rule implementation, unit tests, docs, and registration.
Socratic code review using probing questions instead of direct critique. Based on Feynman's principle that teaching reveals gaps in understanding. Helps developers articulate reasoning, surface hidden assumptions, and discover issues themselves.
Emulated Stripe API for local development and testing. Use when the user needs to process payments locally, test checkout flows, create customers, manage products and prices, handle payment intents, work with webhooks, or use the Stripe SDK without hitting real Stripe servers. Triggers include "Stripe API", "emulate Stripe", "test payments locally", "checkout flow", "payment intent", "Stripe webhook", "Stripe SDK", "STRIPE_API_KEY", or any task requiring a local Stripe API.
Use when reviewing, reproducing, or proving OpenClaw Telegram behavior with a real Telegram user on Crabbox, including PR review workflows that need an agent-controlled Telegram Desktop recording, TDLib user-driver commands, Convex-leased credentials, WebVNC observation, and motion-trimmed artifacts.
Review a code or instruction diff across correctness, simplification, tests, documentation, style, intent, design/API, and performance; add coherence review for metadata changes. Use when the user requests a thorough review before a commit or PR. Validate tool observations and reviewer claims in context, preserve unresolved findings, and report coverage limits. Missing applicable analyzer evidence blocks a verdict
Audits a codebase for removable legacy, unused, duplicate, stale, confusing, or overcomplicated pieces before cleanup. Use when the user asks to simplify a repo, remove dead code, find legacy or unused things, fix duplicate wording/files, or research cleanup opportunities; use codebase-design for module-depth checks and improve-codebase-architecture when cleanup becomes an architecture candidate.
Restructure tangled code into a clear equivalent while preserving behavior. Trigger when the user asks for “readable,” “maintainable,” or “clean” code, a review flags comprehension issues, or the task involves legacy modernization or onboarding cleanup.
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.