Skills
97,501 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Single entry point to install, update, delete, or check the status of any supported CodeArts skill/tool installer. Consolidates four installers behind one skill: superpowers, office-mcp, playwright-cli, and openspec. OpenSpec and other tools don't natively support CodeArts; this meta-installer bridges them. Specify the target via --target <name> (or positionally). Use whenever the user wants to install/update/delete any of these targets, or to list what's supported. Triggers on: install superpowers, install office-mcp, install playwright-cli, install openspec, 一键安装技能, 安装技能, skill installer, update skills, uninstall skills, 安装openspec, 安装office文档工具, 安装浏览器自动化工具.
This skill should be used after creating or modifying skills to update the main README.md file. It scans all skills in ~/.claude/skills/, extracts metadata from SKILL.md files, and regenerates the README with categorized skill listings. Triggers on requests mentioning "update skills readme", "refresh skills list", or after adding new skills.
Create and present web-based slide decks for developers using Slidev with Markdown, Vue components, code highlighting, animations, and interactive features. Use for technical presentations, conference talks, code walkthroughs, teaching materials, or developer decks
Write paid social ad copy variant sets for Meta (Facebook and Instagram), LinkedIn, X, and TikTok placements. Use when the user says "write ad copy", asks for "Facebook ad variants", wants a "LinkedIn ad for this offer", or asks for "ad angles for" a product or promotion. For each requested placement it produces 3–5 variants built from an angle matrix — pain, aspiration, social proof, objection-flip, honest urgency — with every copy part written to the platform's real character limits, plus a one-line creative brief per variant for the designer. It reads social-context.md for brand voice, audience, and offer positioning before drafting. Copy is checked against platform ad policies: no calling out sensitive personal attributes, no fabricated scarcity or invented social proof.
Audit a social profile or content history the user pastes — LinkedIn, X, Instagram, Facebook, Threads, or Bluesky — against their stated goals. Use when the user says "audit my LinkedIn", asks "why isn't this growing", says "review my profile", or asks "what's wrong with my content". It scores posting consistency against intended cadence, pillar balance across recent posts, hook quality with verbatim rewrites, format mix, engagement patterns the user reports, and bio/profile completeness. It reads social-context.md for the user's goals, pillars, and stated cadence so the audit measures against their own intentions rather than generic benchmarks. It works entirely from pasted data — stating plainly what it cannot know without analytics exports and asking the user to paste what they have — and it ends in a prioritized gap list with a 3-move action plan, biggest lever first.
Develop comprehensive social media strategies across multiple platforms. Create content calendars, establish posting schedules, design engagement tactics, plan influencer outreach, and structure paid social campaigns with ROI measurement.
Apply SOLID principles to write flexible, maintainable, and testable code. Use when designing classes, interfaces, and module boundaries. Covers Single Responsibility, Open/Closed, Liskov Substitution, Interface Segregation, and Dependency Inversion with practical TypeScript examples and detection heuristics.
After an agent answers, trace the original basis and sources of that answer. Use when the user says "查出处", "show sources", "原始依据在哪", "这条回答的来源是什么", or wants to verify the URLs/documents that support an answer; first list the sources the agent actually read in the conversation, then check primary sources for claims that lack citations, and write it up as a Markdown file.
When the user wants to plan or execute SOX 404 control testing, select audit samples, build testing workpapers, or classify control deficiencies. Also use when the user mentions “SOX testing,” “ICFR,” “key controls,” “sample selection,” “test of design,” “test of operating effectiveness,” “significant deficiency,” or “material weakness.”
Video voiceover and music. Three tasks: (1) Voiceover: generate narration audio from a script using TTS; (2) Music: match suitable BGM to the video's content and rhythm, find it in a royalty-free music library, and download it; (3) Mixing: combine BGM, narration, and video with ffmpeg, automatically duck the music under speech, normalize loudness, add fades, and align to video length. Trigger phrases include “voiceover,” “music,” “add BGM,” “background music,” “find music,” “add narration to a video,” “TTS voiceover,” “mixing,” and “audio.”
Motion design director and B-roll generator. Turn text, a script, or a video into a 30+ second motion-design B-roll video in one go (HTML → deterministic MP4). No slideshow-style transitions; it starts with “visual metaphor + motion.” Includes multiple design styles (dark SaaS magic, black-and-white typewriter, warm editorial, minimalist white, blueprint grid). Outputs never use emoji, instead using typography, geometry, and linear SVG icons. Trigger phrases include “make B-roll,” “make a motion video,” “motion intro,” “text animation,” “concept animation,” “HTML to video,” “add visuals to this section,” and “30-second b-roll.”
Video cover and thumbnail design. Create high-click-through covers for talking-head videos, Bilibili, WeChat Channels, Xiaohongshu, and YouTube while maintaining the account's consistent visual identity (a unified layout style, with each cover redesigned rather than locked to a template). Trigger phrases include “make a cover,” “video thumbnail,” “title image,” “Bilibili cover,” “Xiaohongshu cover,” “cover,” and “create a cover.” Input: a title, topic, or screenshot; output: a cover image in the aspect ratio for the selected platform.
Talking-head video editing in two stages: (1) Edit the talking-head footage: transcribe, identify verbal slips, filler words, silence, repetitions, stumbles, and sentence fragments; generate a review page; after user confirmation, cut a clean video, then transcribe again and proofread the subtitles with AI. (2) Assemble the final video: combine the edited video, subtitles, and assets into a storyboard, preview the timeline, and export the final MP4. Trigger phrases include “edit this talking-head video,” “remove verbal slips,” “process screen-recording footage,” “finish the talking-head video,” “make a timeline,” “export a vertical video,” and “videocut.”
End-to-end video subtitling—from speech transcription and speech-style segmentation to AI proofreading and correction (proper nouns, homophones, punctuation), generating SRT/ASS, and optionally burning subtitles into the video. Trigger when the user says “add subtitles,” “transcribe this video,” “generate srt,” “proofread subtitles,” “segment,” “burn in subtitles,” “hard subtitles,” or “subtitle.” Input is video or audio; output is a proofread subtitle file, optionally with a finished video with subtitles.
Conversational design workshop for substantial work. Interviews the human one question at a time, explores 2-3 approaches with trade-offs, and presents the design section by section for approval before writing only design.md, then stops. Combines requirements discovery with codebase research and architecture design. Use when the user explicitly requests a spec or when atelier-orchestrator selects a Spec-backed Plan. Ambiguous design or discovery requests route through atelier-orchestrator.
Enforces spec-before-code workflow for AI-driven development. Automatically selects Spec-Kit or OpenSpec mode, triages complexity (quick/standard/thorough), recovers session context, and applies quality gates (G0-G4) with inline self-review at every stage. Use this skill whenever the user says "/super-spec", "spec first", "规范先行", or starts any feature, bugfix, or refactor — especially in projects with .spec-mode, .specify/, or openspec/ directories. Even if the user doesn't explicitly ask for spec-driven workflow, activate this skill for any non-trivial code change to prevent skipping the design phase. Orchestrates: Spec-Kit (v0.7.1, Workflow Engine) / OpenSpec (OPSX v1.2.0) + planning-with-files (v2.30.0) + ui-ux-pro-max (v2.5.0, 67 styles, 161 palettes, 14 stacks, 6 specialist skills) + Superpowers (v5.0.7, inline self-review, subagent model selection) + MemPalace (v3.3.0, 29 MCP tools, cross-session memory, knowledge graph).
Development workflow generator — Generate issue-to-pr-workflow.md for your project. Creates a project-specific development workflow through interactive dialogue, covering branch strategy, quality gates, development style, and optional sub-agent definitions. English triggers: "Generate workflow", "Create development workflow", "Setup issue-to-PR flow" 日本語トリガー: 「ワークフローを生成」「開発フローを作成」「Issue-to-PRフローを設定」
Use when a large uncommitted diff has piled up on a branch and needs to be split into a chain of small, safe, sequential PRs — explicit triggers include "разбей PR", "слишком большой PR", "раздели изменения на несколько PR", "split this PR", "too many files in one PR". Retroactive only: it acts on a diff that already exists, never on work not yet started. Splits the diff into stacked branches/PRs, each capped at 15 business-logic files (assets/docs excluded from the cap) unless breaking the cap is the only way to keep coupled files together. Does not run its own build/lint/test — that's left to the project's own pre-push hook. Tracks progress in a JSON state file (deleted automatically once the split completes) so the split survives session restarts and no file from the original diff is ever silently dropped.
Spring for Apache Kafka integration with KafkaTemplate and @KafkaListener. Covers producers, consumers, retry topics, DLT, and transactions. USE WHEN: user mentions "spring kafka", "KafkaTemplate", "@KafkaListener", "kafka producer Spring", "kafka consumer Spring", "retry topic" DO NOT USE FOR: raw Kafka - use `kafka` skill, RabbitMQ - use `spring-amqp` instead
Spring Data R2DBC for reactive database access in Spring Boot 3.x. Covers R2dbcRepository, DatabaseClient, reactive transactions, and WebFlux integration. USE WHEN: user mentions "r2dbc", "reactive database", "R2dbcRepository", "DatabaseClient", "reactive SQL", "WebFlux database", "non-blocking database" DO NOT USE FOR: blocking JDBC - use `spring-data-jdbc` or `spring-data-jpa` instead, MongoDB reactive - use `spring-data-mongodb` with reactive repository
Design and audit Spring Security configurations for Kotlin plus Spring services, including filter chains, JWT or OAuth2 resource server setup, method security, CORS, CSRF rationale, and public endpoint exposure. Use when adding or reviewing authentication and authorization, narrowing access rules, validating token handling, or checking for insecure defaults and accidental exposure
McKinsey-style storyline framework for building presentation decks. Use when users need to structure presentations, pitch decks, or strategic communications. Creates logical flow where each storyline becomes a slide title, progressing from problem to solution.
Use when an organization needs strategic clarity — identifying where to play, how to win, and whether the position is defensible. Composes Wardley evolution mapping, Lafley's choice cascade, Helmer's 7 Powers, and Working Backwards based on what's already decided.
Guides development with supastarter for Next.js only (not Vue/Nuxt): tech stack, setup, configuration, database (Prisma), API (Hono/oRPC), auth (Better Auth), organizations, payments (Stripe), AI, customization, storage, mailing, i18n, SEO, deployment, background tasks, analytics, monitoring, E2E. Use when building or modifying supastarter Next.js apps, adding features, or when the user mentions supastarter Next.js, Prisma, oRPC, Better Auth, or related Next.js stack topics.
External recon for software supply-chain attack surface — package-namespace squatting candidates, dependency-confusion vulnerabilities, GitHub Actions injection openings, container image registry exposure, SBOM mining, internal-package-name leakage, and CI/CD configuration exposure. Reconnaissance and identification ONLY — actual package publishing / typosquat attacks are EXTERNAL-OFFENSIVE and require explicit written sign-off because they can affect the entire npm/PyPI ecosystem. Use when the target has a public GitHub org, when their build artifacts/SBOMs are reachable, when their docker images are on Docker Hub/GHCR, or when you find internal package names in their JS bundles.
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.