Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Testing & quality
Interactive QA session where user reports bugs or issues conversationally, and the agent files GitHub issues. Explores the codebase in the background for context and domain language. Use when user wants to report bugs, do QA, file issues conversationally, or mentions "QA session".
Reviews Rails pull requests, focusing on controller/model conventions, migration safety, query performance, and Rails Way compliance. Covers routing, ActiveRecord, security, caching, and background jobs. Use when reviewing existing Rails code for quality, conducting a PR review, or doing a code review on Ruby on Rails (RoR) code.
Test an oh-my-opencode-slim release candidate or bugfix before publishing. Use when validating a packed plugin artifact, release branch, crash fix, OpenCode runtime compatibility, or model-specific smoke test such as OpenCode 1.17.11 message transform regressions.
Reviews Remix v2 loaders and actions for mutations-in-loader, missing validation, leaked server fields, wrong return helpers, v1 useTransition holdovers, and revalidation traps. Use when reviewing loader/action code in a Remix v2 codebase.
This skill should be used when the user asks to "fix my skill" or "audit this skill". Make sure to use this skill whenever the user mentions skill quality, structural issues, broken skills, or skill diagnostics — even if they don't explicitly say "repair-skill". Not for adding features or improving effectiveness — use improve-skill. Not for agents — use repair-agent.
Multi-agent code review without any edits or commits — requested rounds of parallel reviewers, synthesis, default-on reproduction of uncertain material findings, and a final report. Use when the user wants a code review of a diff, branch, path, or PR but does NOT want any fixes applied automatically (no edits, no commits, no staging). Activates the `review-anvil` engine in read-only mode.
(forwward) Performs paranoid code review checking trust boundaries, data integrity, performance, race conditions, error handling, and OWASP security. Triggers on code review, pre-merge checks, security audit, bug hunting, or any request to review, audit, or check code quality.
Use when errors occur deep in execution and you need to trace back to find the original trigger - systematically traces bugs backward through call stack, adding instrumentation when needed, to identify source of invalid data or incorrect behavior
Strict discipline for runtime verification in PDCA workflows: the AI performs verification itself in an environment — real preferred over simulated, subject to determinism and safety — and never defaults to a human checklist. Attempt first, resolve a pluggable provider (project script/skill → shared capability skill), and hand off to a human only at a true hard boundary with a stated reason. Use at any workflow verify/check stage (bug fix or feature) to decide WHERE verification runs, WHO runs it, and WHETHER it is sufficient. Referenced by PDCA hosts via frontmatter dependencies. Triggers — 「环境验证」「真实环境验证」「真机验证」「真机测试」「模拟环境验证」「AI 验证」「验证分层」「验证充分性」 / runtime verification, environment verification, real vs simulated. When a check opens a UI, cite a saved non-empty screenshot; non-UI checks do not. Do NOT alias completion-evidence-discipline (evidence freshness) or runtime-evidence-debug (debug instrumentation).
Prepare Rust crates for publishing with Cargo metadata, versioning, README and license checks, package dry-runs, feature verification, docs.rs readiness, examples, changelog notes, and secret-safe release gates. Use before cargo publish, crate release PRs, or sharing a reusable Rust library.
Optimize Rust libraries with criterion benchmarks, feature-flag matrices, cargo bench workflows, dependency gating, regression baselines, and evidence-driven API performance tradeoffs. Use when adding or reviewing Rust library benchmarks, cargo features, optional dependencies, or performance claims before publishing.
Ruthlessly analyze architectural seams—the interfaces, boundaries, and contracts between system components—to expose coupling problems, abstraction leaks, and design failures. Use when asked to review architecture, analyze coupling, find interface problems, improve module boundaries, audit dependencies, or redesign system structure. Produces uncompromising redesign proposals that prioritize correctness over backwards compatibility.
Design and audit product search, marketplace discovery, recommendation surfaces, filters, facets, relevance signals, zero-results, query understanding, ranking drift, result quality evals, fairness, sponsored placement boundaries, and feedback loops. Use when users need to find content, products, creators, docs, apps, skills, or inventory quickly and trust the results.
Set up versioned git hooks via core.hooksPath (no Husky) with lint-staged (Biome, plus Prettier only for languages Biome leaves uncovered and only if the user wants it), plus typecheck and build. Use when user wants to add pre-commit hooks, commit-time formatting/linting/typechecking, or to replace Husky with git's built-in hooks path.
Review changed code for reuse opportunities, quality issues, and inefficiencies using three parallel review agents, then fix any issues found. Triggers when the user says "simplify", "clean up the code", "review the changes", or after run-plan execution when code quality verification is needed.
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy
Run tests, find failures, fix them, verify. Diff-aware: tests affected behavior in web apps, APIs, CLIs, and libraries. This skill MUST be invoked on the triggers below, and SHOULD be invoked after any code change the user has not yet verified. Triggers on: 跑測試, 測一下, 有沒有壞掉, run tests, test this, 跑一下 tests, 測試, check if broken
Audit test suites (unit, integration, E2E) in any language for correctness and quality. Use when the user wants tests reviewed or improved ("audit my tests", "are my tests good"), test coverage or gaps checked, or test strength measured with mutation testing.
The technical-verification layer of the thumb-first mobile suite — finds and fixes objective platform defects in mobile web apps and PWAs (iOS Safari, Android Chrome). Covers viewport/100vh bugs, safe-area insets, touch targets and tap delay, service workers, web push, PWA manifest, and Core Web Vitals — with file/line defects and fixes, not opinions. Target-aware: full coverage for web/PWA; for native apps (React Native, Expo, SwiftUI, Flutter) it detects the target and hands off to a native skill if one is installed, otherwise states plainly that it does not do native-platform audits. Use for mobile-web/PWA technical audits, safe-area/notch issues, 100vh overflow, service-worker/offline, install/manifest, or mobile performance. Usually run via the /thumb-first umbrella (which also runs the design pass); invoke directly for technical checks only.
Find hardcoded colour, spacing and type values and wrong-tier token references in consuming code. Trigger: find hardcoded values, any hex in the code, are we using tokens correctly, token compliance. Do NOT use for token definitions — token-audit; token file format — schema-validator.
Triage a bug or issue by exploring the codebase to find root cause, then create a GitHub issue with a TDD-based fix plan. Use when user reports a bug, wants to file an issue, mentions "triage", or wants to investigate and plan a fix for a problem.
Android device automation via uiautomator2. Use this skill whenever the user wants to click/tap elements, swipe screens, start or stop apps, press hardware keys, type text, take screenshots, inspect UI hierarchy, or perform any actions on an Android device.
Evaluate UI designs and implementations against established UX principles. Use this skill when the user wants a UX heuristic evaluation, usability review, UX critique, or analysis of a design against best practices. Triggers include: 'evaluate the UX', 'heuristic evaluation', 'UX review', 'usability critique', 'does this follow UX best practices', 'what UX issues do you see', 'critique this design', 'review this flow', or when the user provides a screenshot or HTML file and asks for UX feedback. Also trigger when users reference specific heuristics like 'visibility of system status', 'Fitts's Law', 'Hick's Law', 'cognitive load', or any of Nielsen's heuristics or Laws of UX by name.
Enforce portable Vitest house style in apps/packages: per-workspace unit + integration projects, tests/ layout, scripts, passWithNoTests, fixtures/setup, unit vs integration boundaries, and a high-signal test policy (integration over unit when that setup exists, never add new setups unasked, no post-hoc unit tests, failure modes first). Use when scaffolding tests, reviewing test layout, adding Vitest to a package, or when the user asks for vitest-architecture. Optional with-* overlays for testcontainers, Effect testing, Elysia Eden treaty, React happy-dom, typecheck projects, Storybook browser, live harnesses, and coverage. Not a Vitest API docs skill.
Route web-game auto-playtesting with WAI Play (waiterve/wai-play): decide whether the next move is a testability check, authoring or repairing the `GameFlowAgentAPI` bridge, running a real browser playtest, reading the five-dimension quality report, or unblocking a key node the agent cannot reach. Use when the user wants an AI agent to actually play their HTML5 / canvas / vibe-coded web game and return reproducible evidence, scores, and fix suggestions across the five supported types (survivor-like, arcade shooter, platformer, puzzle/card, visual novel). Triggers on: wai-play, WAI Play, auto-playtest, AI plays my game, web game testing agent, GameFlowAgentAPI, GameFlowIntegration, jumpToScenario, game quality score, playtest evidence. Route Unity/Unreal frame-time work to `game-performance-profiler`, engine build failures to `game-build-log-triage`, human playtest notes to `game-demo-feedback-triage`, and generic browser automation to `browser-harness`.
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.