Skills
98,067 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Guide for using Microsoft MarkItDown - a Python utility for converting files to Markdown. Use when converting PDF, Word, PowerPoint, Excel, images, audio, HTML, CSV, JSON, XML, ZIP, YouTube URLs, EPubs, Jupyter notebooks, RSS feeds, or Wikipedia pages to Markdown format. Also use for document processing pipelines, LLM preprocessing, or text extraction tasks.
Use when a project keeps its state in Superself (a `<!-- superself:begin` block in AGENTS.md or CLAUDE.md, or `self setup` resolves the directory to a registered project): read `self context` at session start, attach work to a work unit, report with evidence, and record confirmed decisions so the next session picks up where this one left off.
Generate images with Pruna P-Image models via inference.sh CLI. Models: P-Image, P-Image-LoRA, P-Image-Edit, P-Image-Edit-LoRA. Capabilities: text-to-image, image editing, LoRA styles, multi-image compositing, fast inference. Pruna optimizes models for speed without quality loss. Triggers: pruna, p-image, pruna image, fast image generation, optimized flux, pruna ai, p image, fast ai image, economic image generation, cheap image generation
Automatically review an academic paper using the NeurIPS review form with three reviewer personas, ensemble scoring, and reflection refinement. Extracts text from PDF, runs structured review, and outputs actionable feedback. Use when the user wants to review a paper before submission or get feedback on a draft.
Generate talking head avatar videos with Pruna P-Video-Avatar via inference.sh CLI. Turn a portrait image into a realistic speaking video with built-in TTS. 18x faster and 6x cheaper than competitors. Models: P-Video-Avatar, P-Image (for portrait generation). Capabilities: text-to-avatar, audio-driven avatars, 30 voices, 10 languages, 720p/1080p, built-in TTS, dynamic backgrounds, full-body control. Use for: AI presenters, product demos, explainer videos, virtual influencers, marketing, education, multilingual content, UGC, gaming avatars. Triggers: avatar video, talking head, ai avatar, p-video-avatar, pruna avatar, video avatar, ai presenter, digital human, virtual presenter, lipsync, talking avatar, ai spokesperson, heygen alternative, synthesia alternative, veed alternative, fabric alternative, omnihuman alternative
Use this when reviewing AEM Edge Delivery Services (EDS, Franklin, Helix) code, either self-review at the end of development before opening a PR, or reviewing an existing pull request. Validates block structure, CSS and JS patterns, DOM output, Lighthouse performance, and accessibility against EDS best practices, and posts findings as review comments or GitHub suggestions.
Recommend the best Qwen model and parameters. TRIGGER when: choosing between Qwen models, comparing Qwen model pricing, understanding Qwen model capabilities, asking which models are available on Token Plan, viewing cost history, when an execution skill needs model selection advice, or user explicitly invokes this skill by name (e.g. use qwencloud-model-selector). DO NOT TRIGGER when: non-Qwen model discussions (OpenAI, Gemini, etc.), general AI questions unrelated to Qwen.
The front door to the Small Business plugin. Listens to what the owner needs right now — vague or specific — and routes them to the best skill or slash command for the moment. Also serves as a guide: explains what's available, suggests what to try next, and adapts recommendations based on stored business context. Trigger whenever the owner asks "what can you do," "help me with my business," "what should I focus on," "I don't know where to start," or any open-ended business request that doesn't clearly match a single skill
Use inside a product repository when the user wants the full MVP built from their BuilderOS spec documents. Triggers on phrases like "build my MVP", "build the app", "execute the roadmap", "start the build", "work through the whole roadmap", "build everything", or any request to implement the entire plan rather than a single task or phase. Requires `docs/prd.md` and `docs/product-roadmap.md` (plus `docs/product-vision.md` and `docs/design.md` for context). Works through every roadmap task in order — implementing, testing, and verifying each before moving on, marking checkboxes and updating the status line — and runs until all tasks are complete and the magic moment works end to end, then initializes git with an initial commit and offers to connect a remote repo.
End-to-end testing patterns with Playwright for full-stack Python/React applications. Use when writing E2E tests for complete user workflows (login, CRUD, navigation), critical path regression tests, or cross-browser validation. Covers test structure, page object model, selector strategy (data-testid > role > label), wait strategies, auth state reuse, test data management, and CI integration. Does NOT cover unit tests or component tests (use pytest-patterns or react-testing-patterns).
When the user wants to set up, optimize, or manage Google Ads campaigns. Also use when the user mentions "Google Ads," "Google Search Ads," "PPC," "SEM," "PMF testing with ads," "test product-market fit," "Responsive Search Ads," "RSA," "Performance Max," "Quality Score," "keyword bidding," or "Google Display/YouTube ads." For paid mix, use paid-ads-strategy.
Produce a multi-quarter QA strategy document. Covers scope, risk-based prioritization, test levels (unit/integration/E2E), pyramid analysis, entry/exit criteria, quality KPIs, tool selection rationale, CI scaling levers, and timeline planning. Output is an actionable strategy document, not a shelf document. Use when: "test strategy," "QA strategy doc," "testing approach," "QA roadmap," "multi-quarter QA direction." Not for: a single-sprint or single-release plan — use test-planning. Not for: identifying which areas carry the most risk — use risk-based-testing first. Related: risk-based-testing, qa-metrics, release-readiness, test-planning, test-reliability.
Edit videos locally using ffmpeg. Trim, concat, resize, speed, overlay, extract audio, compress, and convert. Use when: (1) Trimming or cutting video segments, (2) Concatenating multiple clips, (3) Resizing video for social platforms, (4) Extracting or replacing audio, (5) Compressing video, (6) Converting video formats, (7) Getting video info.
Use this when you need more information about an AEM Edge Delivery Services (EDS, Franklin, Helix) feature or guidance on how to implement one, and your existing web-search tools are not turning up relevant results. Searches the aem.live documentation, ranks matching pages by relevance, fetches full articles, and surfaces code examples and deprecation warnings.
Cosmos-Embed1 video-text embedding for text-to-video retrieval, video-to-video search, semantic deduplication, and fine-tuning. Use when the user asks to "fine-tune Cosmos-Embed1", "run cosmos-embed inference", "export Cosmos-Embed1", "embed videos", or "search videos with text".
The mandatory pre-launch gate and four-verb execution contract for every TAO workflow or action. Invoke BEFORE launching anything side-effecting — AutoML, train, evaluate, inference, export, TensorRT engine generation, or DEFT/application workflows — on any execution platform. Covers platform selection, credentials, image confirmation, dataset intake, preflight, the launch review, job records, monitoring, and failure/retry classification. Trigger phrases include "train this model", "run AutoML", "launch on SLURM/docker/k8s/brev/virtualenv", "evaluate my checkpoint", "start a TAO job".
Corrects ASR/STT transcription errors — homophones, garbled terms, person-name errors, mixed Chinese/English — with dictionary rules plus Claude's built-in AI, no external API key required. Use to clean meeting notes, lecture transcripts, interviews, or speech-recognition text, even when the user only says “fix this transcript”. Not for structuring a clean transcript into minutes (use meeting-minutes-taker).
Initializes the Claude Code .claude/ structure (CLAUDE.md, Agents, Rules, Skills, hooks) in any target repository. Use for requests such as "set up Claude," "create .claude," "initialize CLAUDE.md," "organize Agents," or "set up claude-code." Use update-claude to enrich an existing .claude/ with differences. Also includes preparing prerequisites for implement-issue-tree (gh auth / sub_issues / workflow js).
Turn references you admire into a scene-by-scene Claude Motion prompt for your own topic and brand. Takes a video for the moves (screenshots, frames or a contact sheet of a reel or ad) and, optionally, a look reference for the style (a website, a brand, a moodboard). Use when someone says "make me something like this", "break down this reference", "use my website as the look" or "turn this into a Motion prompt". Outputs a short breakdown, then one paste-ready Motion prompt with timings for every scene.
Record and cut a short showcase video of a game mod, or a compilation of clips, ready to post on X/YouTube. Records the game window with only the game's own audio, picks moments from contact sheets, and edits with one-line animated titles, creator credits, transitions, music, watermark and fade-out via an EDL JSON. Use when the user wants a demo video, trailer, clip or montage of a mod, or wants to combine existing clips into a styled video.
Distill a book, long-video transcript, podcast, course, or interview into a coherent set of executable skills. Use when the user asks to "拆书" / "蒸馏一本书" / "把 XX 书做成 skill" / "把这个视频/播客/课程蒸馏成 skill" / "turn a book or video into skills" — i.e. wants the frameworks, principles, and methodologies in long-form content extracted into atomic, reusable Claude skills that an agent can invoke in real-world situations. NOT for simple summarization, book reviews, or role-playing as the author (that is nuwa-skill's job).
Use this when implementing code changes in AEM Edge Delivery Services (EDS, Franklin, Helix), whether new or modified blocks, core functionality (scripts.js, styles, delayed.js, etc.), or both. Creates and modifies block folders and decorate functions, updates core scripts, scopes CSS, and wires up delayed loading. For the overall development process use content-driven-development.
Apply an emotional resonance lens to a UI that is technically correct but flat, prescribing changes at the copy, motion, and interaction layer. Use when a design tests fine but lands cold. For the polish-perception argument, use `aesthetic-usability` (ui-design).
Generate videos with Pruna P-Video and WAN models via inference.sh CLI. Models: P-Video, WAN-T2V, WAN-I2V. Capabilities: text-to-video, image-to-video, audio support, 720p/1080p, fast inference. Pruna optimizes models for speed without quality loss. Triggers: pruna video, p-video, pruna ai video, fast video generation, optimized video, wan t2v, wan i2v, economic video generation, cheap video generation, pruna text to video, pruna image to video
Shared Cosmos3 frontend that explicitly routes Cosmos Framework and Cosmos-RL, validates runtime model/video-dataset/SLURM inputs, consumes an SQSH or packaged backend image, optionally plans explicit clean source builds, prepares checkpoints, validates the first update in-process, and returns token-weighted losses and task-aware accuracy.
TypeGPU is type-safe WebGPU in TypeScript. Use whenever the user writes, debugs, or designs TypeGPU code: 'use gpu' shader functions, tgpu.fn, buffers, textures, bind groups, compute and render pipelines, command encoders, render passes, render bundles, vertex layouts, slots, accessors, @typegpu/react hooks (useRoot, useFrame, useUniform), React Native worklet rendering, and any TypeGPU API. Shader logic and CPU-side resources are tightly coupled - handle both sides here even if the user only mentions one (e.g. "how do I write a shader", "how do I create a buffer"). Trigger on any mention of typegpu, tgpu, "use gpu", TypedGPU, or WebGPU code written using TypeGPU's schema API (d.*, tgpu.*, std.*). Do NOT trigger for raw WebGPU (using GPUDevice/GPURenderPipeline directly without tgpu), WGSL-only questions, Three.js, Babylon.js, or WebGL.
Use this when the page-import pipeline needs to survey the block palette available from a local AEM Edge Delivery Services project and the Block Collection to inform content modeling decisions. Scans local block folders, searches the Block Collection for common blocks, resolves each block's purpose, and returns a consolidated block inventory annotated with per-block purpose and availability. Do not invoke directly — called by page-import as a pipeline step
This skill should be used when the user asks to "make a bar chart race", "turn a CSV into a video", "build an animated data visualization", "create an animated statistics video", "animate a chart/graph", "make a number counter/ticker", "animate a ranking over time", or "batch-render one chart template across many datasets". Covers data-driven keyframing, value interpolation, rank transitions, counters, pacing, annotation, and template×data batch output.
Go-to-market strategy for AI products. Use when positioning AI products, handling "who is responsible when it breaks" objections, pricing variable-cost AI, choosing between copilot/agent/teammate framing, or selling autonomous tools into enterprises.
Help users plan products and strategy when outcomes are unpredictable. Use when someone is dealing with ambiguous timelines, building in fast-moving markets, planning AI/ML projects, or asking how to make commitments when they don't know what will happen.
Use this for any AEM Edge Delivery Services (EDS, Franklin, Helix) code change—new blocks, block modifications, CSS styling, bug fixes, core functionality (scripts.js, styles, delayed.js), or any JavaScript/CSS work that needs validation. Covers the Content Driven Development process: identify or create real test content before coding, design an author-friendly content model, and validate against that content throughout development. Do not use for documentation-only changes or configuration changes that do not affect authoring.
This skill should be used when the user asks to "triage a vulnerability report", "assess a CVE", "evaluate a bug bounty submission", "decide if a finding is valid", "review a security finding", "dismiss a vulnerability", "should we fix this CVE", "prioritize a vulnerability report", or needs to determine whether an incoming vulnerability report warrants investigation. Applies 7 brocards (rules of thumb) to systematically accept, dismiss, or request more information on vulnerability reports, or needs to filter raw findings from agentic vulnerability discovery pipelines before human review.
Read-only storage analysis assistant for macOS / Windows (automatically detects the system). Scans disk usage across the entire computer, identifies what is taking up space, classifies each item into three levels—🟢 safe to clean automatically / 🟡 requires human judgment / 🔴 clean with caution—and provides actionable recommendations. Generates a polished, collapsible interactive HTML report with one-click command copying, and can launch a local service to delete items from a web page with one click (move to Trash/delete permanently). Scanning is read-only throughout. Always use in these situations: the user says “storage analysis,” “disk full,” “C drive/hard drive full,” “not enough space,” “free up space,” “clean disk,” “taking up space,” “what’s taking up space,” “help me check storage,” “check computer storage/space,” “storage space,” “computer is low on space,” “memory full/not enough/insufficient,” “check memory/storage” (in colloquial Chinese, “memory” often means storage), “disk cleanup,” “clear cache,” or “磁盘清理”; or complains that the computer is out of space, wants to know what is using the disk, or wants cleanup recommendations. Note: if the user explicitly means RAM (e.g. “which process is using memory,” “high memory usage,” or wants to view Activity Monitor), that is RAM, not storage, and is outside this skill.
Pick, install, and wire up Capacitor plugins for native features (camera, filesystem, geolocation, biometrics, in-app purchases, social login, SQLite, calendar, contacts, health, BLE, NFC, etc.). Covers official @capacitor/* packages (Capacitor 8.5 stable, 9 alpha on next) plus the 145-package @capgo/* index, native permission keys (Info.plist, AndroidManifest, variables.gradle minSdk), npx cap sync / cap ls verification, and errors like '"Camera" plugin is not implemented on ios', 'unable to find plugin', UNIMPLEMENTED, and OS-PLUG-* codes. Use for "which plugin should I use for X", adding a plugin, or a plugin call failing after install. Do not use for push setup (capacitor-push-notifications), splash assets (capacitor-splash-screen), keyboard layout (capacitor-keyboard), deep links (capacitor-deep-linking), OTA updates (capgo-live-updates), upgrading plugin source code (capacitor-plugin-upgrades), or adding SPM to a plugin (capacitor-plugin-spm-support).
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.