Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Creative & media
Author, load, and troubleshoot Reality Composer Pro Shader Graph materials for RealityKit on visionOS. Use when building Shader Graph materials, exposing promoted inputs for runtime control, debugging exported USD and MaterialX interop, or building shader graphs programmatically in Swift on visionOS 27.
Use the Drawcall Speech API from apps, websites, and 3D experiences. Use when explaining or showcasing text-to-speech usage with https://v1.speech.drawcall.ai, including simple HTML audio, JavaScript audio loading, voice selection, style control, browser autoplay constraints, URL encoding, caching behavior, and first-request warmup.
Case studies of television series writing (剧集案例库) from primary texts; Sorkin's six West Wing teleplays (act-page tables, eight act-out types); Chase's five Sopranos scripts (the 35-beat A/B/C/D recipe, what a pilot refuses to explain); Fellowes's annotated Downton Abbey season two (19 storylines, 419 footnotes); Waller-Bridge's Fleabag Scriptures (the six-episode half-hour template); Calvisi's eight minute-by-minute pilot beat sheets (Breaking Bad, Mad Men, Game of Thrones and more); Landau's 47-series structure appendix; Miller's Hannibal and HTGAWM breakdowns; Goldberg & Rabkin's beat sheets; Sakamoto Yuji's We Made a Beautiful Bouquet (a feature) and Noh Hee-kyung's The Most Beautiful Goodbye. Use when you want a worked example of how an episode or pilot is built, need real act lengths, scene counts or storyline grids to imitate, are breaking a multi-thread ensemble episode, designing a bottle episode or a finale, writing a relationship or family drama, or ask 某一集是怎么写的 / 这部剧的结构表.
Validate advanced Three.js graphics as authored systems rather than subjective screenshots. Use for fixed-view visual contracts, field and pass diagnostics, no-post baselines, seed sweeps, camera-scale tests, temporal stability checks, GPU budgets, and regression evidence for procedural scenes.
Use meitu-cli to remove video grain, color blocks, and motion noise, reducing noise while preserving detail. Triggered when the user mentions video denoising, noise reduction, degraining, reducing noise, too much noise, heavy grain, color artifacts, or high ISO.
Uses meitu-cli to remove passers-by, subtitles, watermarks or logos from videos, or to clear all on-screen text with one click. Triggers when the user mentions removing passers-by, removing pedestrians, removing subtitles, removing dialogue, removing watermarks, removing logos, removing corner badges, erasing a Douyin ID, clearing the screen, a clean picture, or "no text should appear".
Use meitu-cli for video motion and pose transfer, transferring the motion from a reference video onto a person in an image while preserving their appearance. Trigger only when the user explicitly provides a person image and a motion reference video and expresses motion transfer, pose-driven animation, imitating a motion, or making the person in the image follow the video's motion.
Build image-model-driven PPT decks from a user topic, source materials, papers, reports, notes, datasets, existing slides, or style references. Use when the user asks to create, redesign, package, or QA a PowerPoint deck where every slide should first be generated as a high-quality full-slide image and packaged into PPTX with manifests, previews, and evidence-backed QA. Layered editable reconstruction is optional and only used when explicitly requested. This is the general deck-building entry. Prefer narrower skills only when the request is specifically an academic/research PPT workflow, a prompt/method clinic, a sci-employee course/video wrapper, or a recurring Chronicle research/script loop.
Generate text-to-video with Wan 2.7 (Wan-AI's flagship motion model) on RunComfy. Documents Wan 2.7's strengths (multi-reference conditioning, audio-driven lip-sync via `audio_url`, smoother transitions, prompt expansion), the duration / resolution / aspect-ratio schema, and when to route to HappyHorse 1.0 / Seedance 2.0 / Kling / LTX 2 instead. Calls `runcomfy run wan-ai/wan-2-7/text-to-video` through the local RunComfy CLI. Triggers on "wan", "wan 2.7", "wan-2-7", "wan video", or any explicit ask to generate video with this model
Run an end-to-end webtoon production harness — a 27-agent / 4-phase Claude Code agent team that takes a single episode from trend research to a finished vertical-scroll viewer. Use when the user wants to produce a webtoon episode: research popular webtoon trends, write a dialogue-heavy, high-tension, twist-every-episode scenario, render character reference sheets first, render 50+ panels per episode with in-image speech-bubble (Korean dialogue) baking via codex-image, run a generation–validation loop, and assemble a vertical-scroll viewer. Triggers on: webtoon harness, make a webtoon, produce a webtoon episode, webtoon scenario to image, next episode, redraw panel, stronger twist, webtoon trend research. Web data gathering for the trend-research phase routes through the `scrapling` skill. Pure webtoon recommendation/critique is answered directly.
Make one video accessibility-ready: captions on, on-screen text sized and contrasted to be readable, pacing checked so every step can be followed. Use when the user says "accessibility pass", "make this video accessible", "check WCAG for my video", "is this readable for everyone", or "prep this for an accessibility review".
Annotate an existing video with callout boxes and arrows that point at the exact buttons and fields the narration mentions, appearing and leaving in time with the voice. Use when the user says "add callouts", "add arrows", "annotate my video", "point at the buttons", "label the steps on screen", or "highlight where to click".
Put the takeaway on screen: add short text overlays that surface the key point of each section of an existing video as it's narrated. Use when the user says "add key points on screen", "add takeaway text", "put the main points as text overlays", "reinforce the message visually", or "viewers skim - make the points readable".
Replace a video's cold open with a 5-second hook that states the payoff or problem up front, before the walkthrough starts. Use when the user says "add a hook", "the opening is boring", "it takes too long to get to the point", "give it a stronger start", "lead with the payoff", or shares a video that opens straight into a UI with no reason to keep watching.
Generate AI video from any input — text, image, or script — with Pexo. Auto-routes each shot to the best of 10+ models (Seedance, Kling, Veo, Sora, and more), writes the prompts, generates every shot, and returns a finished video with music and subtitles. Use for any AI video generation request: "ai video generation", "generate a video with ai", "ai video generator", "make an ai video". NOT for editing an existing video.
Use when creating cloned voices with Alibaba Cloud Model Studio CosyVoice customization models, especially cosyvoice-v3.5-plus or cosyvoice-v3.5-flash, from reference audio and then reusing the returned voice_id in later TTS calls.
Operate AudioShake's cloud source-separation service (developer.audioshake.ai) to split a recording into stems. Use when a task involves isolating vocals, drums, bass, guitar, piano, keys, strings, or winds from music; separating dialogue / music / effects (DME) for film-TV post-production, localization, and dubbing; splitting a mixed recording into one stem per speaker; lyric transcription or word-level alignment; music detection/identification; or speech denoise/dereverb. Covers the Tasks API job lifecycle (assets, targets, formats, polling vs webhooks, credits, limits), choosing which stem targets a production actually needs, reviewing separated stems for bleed / artifacts / transient smearing / phase problems and repairing them, the rights and consent questions raised by separating copyrighted or third-party recordings, and the decision boundary for when a local open model such as Demucs is the better route than the API. Not for generating or synthesizing new audio, mixing, or mastering.
Physics in Blender 5.2 — rigid bodies (density mass, shapes, bake to keys), classic cloth presets and pinning, new node-based Cloth/Hair Dynamics (XPBD, Collider, effectors), Mantaflow fluids, hair curves, GN simulation zones, and headless baking. Use for destruction, drapes, flags, falling objects, liquids, smoke, and hair motion
Make short hand-drawn 2D animations entirely in JavaScript — storybook / textured-editorial illustration (paper grain, hatching, blue construction guides, glossy insect characters, nest cross-sections) and brush-ink + watercolour explainers — rendered frame-exact with node canvas and ffmpeg, with synthesised sound. Use for "bikin animasi", "animate this story", "hand-drawn animation", "storybook animation", "JS animation", "line drawing animation", "animasi semut", kids' story or science explainer animations. Not for UI micro-interactions, slide decks, or talking-head video editing.
Create and refine image files from Claude Code using a Codex subscription. Handles short creative interviews, distinct visual directions, reference-based edits, consistent asset sets, and verified transparent PNGs. Use for image generation, imagegen, GPT Image, hero art, product photos, icons, posters, 이미지 생성, 시안, 이미지 수정, or 투명 배경. Exclude image analysis and code-native SVG edits.
Codex Pet generator on RunComfy. Build a Codex-compatible Codex Pet spritesheet.webp + pet.json from a single reference image, drop it into `${CODEX_HOME:-$HOME/.codex}/pets/<name>/` and Codex picks it up as a custom Codex Pet next to the 8 built-ins. This skill produces the exact Codex Pet atlas Codex expects (1536x1872 PNG/WebP, 8 cols x 9 rows, 192x208 cells, 9 animation states — idle, running-right, running-left, waving, jumping, failed, waiting, running, review). Calls OpenAI GPT Image 2 edit ONCE via the local RunComfy CLI as `runcomfy run openai/gpt-image-2/edit` to produce a canonical Codex Pet pose, then assembles all 9 animation rows programmatically with ImageMagick micro-transforms — no Codex Pro, no `$imagegen`, no OPENAI_API_KEY required, only RUNCOMFY_TOKEN. Triggers on "codex pet", "create codex pet", "make codex pet", "hatch codex pet", "/hatch image", "desktop pet codex", "codex pets", "spritesheet.webp", or any explicit ask to build a custom pet for OpenAI Codex.
Turn one master product demo into N persona-targeted variants: duplicate the demo per persona, re-script the narration around each persona's pains and outcomes, reorder and trim so every variant leads with what that audience cares about, and close with a persona-specific CTA. Use when the user says "make a version of this demo for each persona", "tailor this demo for admins vs execs", "persona variants of my demo video", or "one demo, different audiences".
Generate full songs and instrumental tracks with ElevenLabs Music on RunComfy via the `runcomfy` CLI. ElevenLabs Music turns a style description plus structured lyrics into studio-quality 44.1 kHz stereo audio — 5 seconds to 5 minutes — with section-level control (Intro / Verse / Chorus / Bridge), multilingual vocals, and commercial-friendly output. Generate a backing track, a full vocal song, a jingle, a podcast intro, a game loop, or an instrumental bed. Calls `runcomfy run elevenlabs/elevenlabs/music-generation` through the local RunComfy CLI. Triggers on "generate music", "make a song", "AI music", "background music", "instrumental track", "ElevenLabs Music", "soundtrack", "jingle", "theme music", "royalty-free music", "compose", or any explicit ask to generate music or a song from a text description.
Turn any complex topic or question into a short, plain-language explainer video: a hook, an analogy, a few essential parts, and a payoff, carried almost entirely by kinetic typography and keyframed shapes since there's usually no product or screen recording to lean on. Use when the user says "explain X simply", "make a video explaining how X works", "turn this concept into a video", or asks for a short explainer on something with no existing footage or article behind it.
Apply, configure, and modulate WebGPU post-processing shaders, stylized looks (pixel art, ASCII, VHS, comic, watercolor, risograph, holographic — and which look fits which request), cinematic color grading, tone mapping, procedural VFX, and on-device vision (object tracking, instance segmentation, pose, person mattes) in framefields. Use when adding blur, grain, LUTs, relighting, curves, greenscreen color keying, subject cutouts, or tracking-driven effects in framefields compositions.
Turn a feature-walkthrough recording into a short, sound-off-safe clip meant to be embedded inside the product itself: a tooltip, onboarding checklist step, or help-widget video, not a standalone shareable video. Trims to the single feature the widget sits next to, adds on-screen text so the point lands muted, and reflows to a compact aspect ratio if the embedding surface needs one. Clueso produces the exported clip; embedding it into the guidance tool is the user's own step. Use when the user says "make an in-app guidance clip for this feature", "turn this walkthrough into an embeddable tooltip video", "create a short in-product help clip", or "I need a muted clip for our onboarding checklist widget".
De-AI-ify rewriting assistant. Triggers when the user says "去 AI 味", "帮我改得不像 AI", "这段太 AI 了", "humanize", "降 AI 率", "改得像真人写的", "去机器味", "文案太生硬了", or "读起来像 AI 写的". Also applies when the user says "帮我改稿" or "润色一下" and the context implies the text shows signs of AI generation. Scans the text sentence by sentence for AI writing traces, rewrites it into a human style through a six-step process, and outputs a comparison report plus the rewritten full text.
Make a showreel-grade product film (landing-page loop, launch video, promo, demo reel, social cut) in code with Remotion, built on the product's own design system, components and voice. Use when someone asks for a product video, landing video, promo, launch film, explainer or "motion design video" for their app, SaaS or codebase. It asks what to include, then covers design discovery, story, music sync, transitions, review loops and a verified final render.
Strip ums, ahs, false starts, and rambling from a video's narration - clean the transcript, regenerate the voiceover, and keep the visuals aligned to the new timing. Use when the user says "remove the filler words", "clean up the ums and ahs", "the narration sounds rambly", "tidy up the voiceover", or shares a video whose spoken track is full of hesitations and false starts.
Generate before-and-after transformation video prompts for Seedance 2.0 on Higgsfield. Use for transformation reveals, glow-ups, makeovers, renovation reveals, fitness transformations, design before-after, business growth visuals, or any content showing dramatic change. Triggers on before after, transformation, reveal, glow up, makeover, renovation, redesign, progress, results.
Generate personal brand and founder story video prompts for Seedance 2.0 on Higgsfield. Use for authority content, day-in-the-life, founder story, personal brand building, thought leadership, creator content, lifestyle videos, behind-the-scenes, or any video meant to build a personal brand presence. Triggers on personal brand, founder, creator, authority, lifestyle, behind the scenes, day in the life, thought leader.
End-to-end AI video production skill for agentic frameworks. Use when the user asks to make, edit, repurpose, caption, soundtrack, or export videos using HeyGen avatars, Seedance or ByteDance b-roll, OpenAI image generation and editing, AI UGC ads, Remotion, HyperFrames, screen recordings, FFmpeg, captions, ElevenLabs music, Suno, or social-video variants.
Direct advanced Three.js camera systems. Use for scale-aware chase rigs, thrust lag, side/orbit cameras, body-relative up vectors, quaternion handoffs, authored cinematic framing, floating origins, pointer-look controls, camera collision constraints, projection ownership, and lifecycle restoration.
Add audio to existing videos with meitu-cli: add BGM, add narration dubbing, or BGM + narration in sequence. Triggered when the user mentions adding background music, adding BGM, changing background music, adding narration, adding dubbing, adding a voiceover commentary, adding voice-over, dubbing and music, or adding audio to a video.
Use meitu-cli to increase video frame rate and interpolate frames (e.g. 24fps → 60fps), making motion in low-frame-rate videos smoother. Triggers when the user mentions frame interpolation, adding frames, raising the frame rate, upconverting the frame rate, 60 frames, 120 frames, smoothness, insufficient frame rate, or video stuttering.
Use meitu-cli to concatenate multiple video clips in sequence, supporting plain concatenation that keeps the original audio or concatenation + background music (automatic lookup / custom music URL). Triggered when the user says concatenate videos, merge videos, video compilation, or concatenate with music.
Plan, implement, and review Alibaba Wan video generation using Alibaba Cloud Model Studio/DashScope hosted APIs or official Wan 2.1/2.2 open-weight checkpoints. Use for Wan text-to-video, image-to-video, first/last-frame or continuation, reference-to-video, video editing, speech-driven video, character animation, regional deployment, billing, licensing, and production safety.
Use when the user needs an end-to-end article illustration workflow in this repository that preserves Type x Style planning, loads article-illustration preferences, recommends Alibaba Cloud image backends, and produces a Markdown article with inserted local image references.
Turns limitations into creative fuel by strategically imposing constraints that force novel thinking, break habitual patterns, and reveal unexpected solutions. Covers resource constraints (budget/time/tools), format constraints (length/medium), rule-based constraints, and perspective constraints. Use when brainstorming feels stuck or generates obvious ideas, working with limited resources, designing with specific limitations, or when user mentions "think outside the box", "we're stuck", "same old ideas", "tight constraints", or "limited budget/time".
Detect whether a short-video verbatim script violates the community guidelines of Douyin, Xiaohongshu, or WeShip Channels; identify violation risk points; provide a tiered risk report and concrete revision suggestions. Trigger words: "content compliance check", "verbatim script check", "risk check", "short-video compliance", "platform guideline check"
Use DeepMotion Animate 3D for markerless video-to-3D human motion capture through the cloud portal, sales-gated API, or real-time SDK; covers capture planning, body/hand/face/multi-person limits, custom characters, retargeting, refinement, exports, DCC/game-engine handoff, QA, pricing, privacy, consent, and rights checks.
Produce speech and clone voices with Fish Audio — its hosted TTS API (S2.1-Pro / S2-Pro / S1 model lineup, REST + WebSocket streaming, instant and persistent voice cloning) and its open-weight OpenAudio S1-mini / Fish-Speech models for self-hosting. Use this skill when an agent must generate narration or dialogue through Fish Audio, choose between Fish Audio's hosted models and open weights, clone a voice from reference audio, author emotion/tone/special markers for expressive delivery, estimate cost from UTF-8 bytes, wire real-time streaming for a voice agent, decide whether self-hosting beats the API, or review Fish Audio TTS output for production. Do not use it to pick a different provider — it covers Fish Audio specifically.
Build or optimize high-quality generation prompts from copy, product information, scripts, rough ideas, reference-image descriptions, or existing prompts. Use for image, video, audio, and mixed creative prompts when the user wants ready-to-copy prompt text, not tool execution. Do not use when the user specifically asks to generate an image with GPT Image or asks for storyboard/shot-list planning.
Generate, reference, and conversationally edit short videos with Google's Gemini Omni Flash through the Gemini Developer API or Gemini Enterprise Agent Platform. Use when a task specifically needs Gemini Omni video, multimodal reference roles, time-directed prompting, audio-aware generation, Interactions API state, or a precise comparison with Veo, Nano Banana, consumer Gemini, or gateway routes.
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.