Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Creative & media
A short skit where every beat is laid down to serve one punchline. The joke has to land on something visible, and the absurdity only works when the camera and the physics stay dead serious. Use when the user wants a Seedance 2.5 / 2.0 video prompt for this kind of clip (Twist-ending comedy skit; 中文触发词: 反转结尾搞笑短片, 搞笑, 反转, 梗, 沙雕), or asks how published cases of this type were prompted. Read references/cases.md before drafting.
Use when the user asks to generate, remix, poll, list, download, or delete Sora videos via OpenAI's video API using the bundled CLI (`scripts/sora.py`), including requests like "generate AI video," "Sora," "video remix," "download video/thumbnail/spritesheet," and batch video generation; requires `OPENAI_API_KEY` and Sora API access.
Product photography enhancement for ad creatives using banana-claude image generation. Takes a product image and generates 5 professional photography styles for ad use: Studio, Floating, Ingredient, In Use, and Lifestyle. Requires banana-claude (v1.4.1+) with nanobanana-mcp. Triggers on: product photo, product photography, photoshoot, enhance product image, product shoot, product photos for ads, generate product photos, studio shot, lifestyle photo.
Use when generating reference-based videos with Alibaba Cloud Model Studio Wan R2V models (wan2.6-r2v-flash, wan2.6-r2v). Use when creating multi-shot videos from reference video/image material, preserving character style, or documenting reference-to-video request/response flows.
Generate App Store / Google Play preview videos using Remotion. Scans the target app's codebase, proposes custom scenes, and writes bespoke .tsx scene files using shared animation primitives. Mobile apps only (iOS/Android). Use when the user wants to create a demo video, app preview, promo video, or marketing video for a mobile app. Builds on appshot-core foundation.
Chanjing AI creation Open API client: submit image/video tasks, poll task status, list and get tasks, and optionally download outputs when the user runs download_result.py. Reads and writes local credentials.json; calls open-api.chanjing.cc.
This skill should be used when the user asks to "record a terminal demo", "generate a CLI demo GIF", "make a vhs recording", "create a terminal GIF for the README", or "capture a CLI walkthrough as a GIF". Generates CLI demo GIFs using vhs (Charmbracelet) for README files and documentation.
Chinese-language writing based on a topic or reference article. Automatically searches the material library, loads the primary style profile, and produces articles that match the user's personal style. Supports WeChat Official Accounts, Twitter, Xiaohongshu, blogs and other scenarios. Use this skill when the user says "写一篇关于XX的文章", "围绕XX创作", "基于这篇文章写一篇", "参考这篇英文写个中文版", "帮我写", "出一篇稿子". Consider using it even when the user provides only a topic or a single reference article.
This skill should be used when the user asks to generate content such as titles, slogans, dialogues, or scripts. It provides content generation capabilities for various platforms (WeChat, Xiaohongshu, Zhihu, Douyin) with support for batch generation, history deduplication, and diversity guarantee. Supports podcast script generation with platform-specific adaptation.
Use when the user wants Codex to learn DaVinci Resolve through Blackmagic Design lesson homework, operate the Resolve UI with Computer Use, validate skills against real Resolve behavior, or build a capstone workflow that turns local/T7 media into Instagram and YouTube cuts.
Use when writing or reviewing JavaScript/TypeScript in this repo that calls Deepgram Text-to-Speech v1 (`/v1/speak`) for audio synthesis. Covers one-shot REST via `client.speak.v1.audio.generate` and streaming WebSocket via `client.speak.v1.createConnection()` / `connect()`. Use `deepgram-js-voice-agent` when you need full-duplex STT + LLM + TTS instead of one-way synthesis. Triggers include "TTS", "text to speech", "speak", "aura", "streaming TTS", and "speak.v1".
Automatically illustrate Markdown documents by turning image markers into browser screenshots or AI-generated images, then writing an image-enriched Markdown output. Use when a document needs screenshots, generated visuals, semantic image placement, or end-to-end document illustration automation.
Agent-callable ElevenLabs tools — generate spoken audio from text, create sound effects and multi-speaker dialogue, re-voice and clean up audio, transcribe audio and video, design synthetic voices, and manage voices, history, and quota. Use when the user mentions ElevenLabs or wants AI audio work — text to speech, narration, voiceover, transcription, voice changing, or sound design — even if they don't name ElevenLabs explicitly, e.g. "read this aloud", "make an MP3 of this", "transcribe this recording".
Transcribe audio/video files to text with Gemini — languages auto-detected, optional speaker labels, timestamps and a vocabulary of names/terms. Supports MP4, MP3, WAV, OGG, MKV and other ffmpeg-readable formats; long recordings are chunked automatically
Plan and implement conversions between game meshes, SDFs, voxels, images, contours, and smooth curves while preserving enough gameplay and rendering data. Use when repairing meshes, extracting meshes from fields, voxelizing geometry, contouring, denoising voxel data, generating LODs, or moving geometry between representations.
Generate crisp, perfectly smooth MP4 videos from a web page or animated visualization by driving headless Chromium over the Chrome DevTools Protocol (CDP) with deterministic frame-stepping, then encoding with ffmpeg and (if remote) sharing via a cloudflared tunnel. Use when asked to make/generate a demo video, explainer clip, launch/marketing animation, social teaser, or any short rendered video from a web scene — especially animated stat/timeline/diagram visualizations — and when you need to iterate on it (tweak, re-render, compare). Solves the common "the recording has hitches / judder / dropped frames" problem.
Generate or edit images using the OpenAI Image API (gpt-image-2). Use when the user asks to generate, create, draw, render, illustrate, mock up, or edit an image, icon, logo, mockup, illustration, OG image, blog hero, marketing asset, or similar visual. Also use when the user supplies a reference image and asks to modify, restyle, or remix it. Triggers on: "generate an image", "create an image", "make a picture of", "edit this image", "restyle this", "make a mockup of", "draw a", "render a", "illustration of".
Turn a tweet, a video, a YouTube video or a landing page into a 9:16 reel (Instagram Reels, TikTok, Shorts) that says what the source says and looks like it, with every cut on a word. Read the source (its words, its media, its measured fonts and colours), pick a structure (a promo, a showcase, "found this", an explainer, before and after), write the script in the creator's voice from the source's own facts, make the voice with timestamps (their cloned ElevenLabs voice), then build the picture from HTML scenes drawn from one clock (window.__frame(t)) in the source's look, opening on its cover (the hook's best moment as the first frame), its demo video framed and cut to words, and assemble it in Fil's video editor over MCP. The creator is in it from the first draft as a placeholder in the camera layouts (a backlit studio silhouette) (full frame, a bubble, half the frame), swapped for their phone takes when they record, and everything retimed to their own words by word anchors, t = W(τ); a reel can also be faceless. Use when someone wants a reel, a short or a vertical video from a post, a thread, a launch, a landing page, a demo video or a YouTube video, "a reel like this one", or a talking-head promo cut to their own delivery.
Remove traces of AI generation from Chinese text. Designed for the distinctive problems of Chinese AI writing: era-opening preambles, overused connectives, internet jargon, translationese, overly formal written style, formulaic structure, canned endings, and intrusion of mainland Chinese terms. Makes text more natural, more human, more like something a Taiwanese person wrote. Trigger words: /humanizer, humanize, 去除 AI 痕迹, 人性化, 去機器味
Professional logo design studio that produces 9+ SVG logo concepts through brand discovery, archetype mapping, and iterative refinement, then generates a complete app asset package (iOS, Android, macOS, Windows, favicons, PWA, social) from the final selection and optionally produces a multi-page brand guidelines document (logo, color, typography, layout, UI components, motion, voice, asset management). Use when the user asks for a logo, brand mark, icon, wordmark, app icon, visual identity, or brand guidelines for a business, product, or project.
Integrate Modellix's unified API for AI image, video, and audio workflows. Use this skill whenever the user wants to generate or edit images, create or transform videos, synthesize speech, transcribe audio, clone a voice, do virtual try-on, or call any Modellix model API. Also trigger when the user mentions Modellix, model-as-a-service for media generation, or providers such as Qwen, Wan, Seedream, Seedance, Kling, Hailuo, MiniMax, Whisper, or CosyVoice through a unified API, or when they ask for a Modellix model's request schema, OpenAPI contract, or required input fields. Prefer modellix-cli (model get-schema, model run --wait, task download, doctor, model list) over hand-rolled REST polling whenever the CLI is available.
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
Generate and edit images from the CLI using picture-it. Use this skill whenever the user asks to create, edit, or manipulate images — blog headers, social cards, hero images, product comparisons, YouTube thumbnails, movie posters, magazine covers, Instagram edits, background removal, or any visual content. Also trigger when the user mentions picture-it by name, wants to composite images, apply color grading, add text to images, remove or replace backgrounds, crop/resize photos, or needs any kind of image generation or photo editing from the terminal. This skill covers multi-pass AI image editing workflows that chain composable operations together.
EffectComposer setup and architecture for Three.js post-processing pipelines. Use when setting up multi-pass rendering, combining effects, creating custom passes, managing render targets, or building reusable effect stacks. Foundation skill for all post-processing work.
Transform photos, film stills, or new scene descriptions into the observed visual grammar of the 2026 animated film 《牛来》. Use when the user asks for 牛来风格、牛来正片风格、粗粝低模动物 3D、把真人画面改成牛来式动画, or requests iterative comparison against 《牛来》 references. Default to the rough film-3D mode; use the separate ink-poster mode only when the user explicitly asks for the movie poster look.
Use when the deliverable is spoken words to be performed on camera or into a mic: video and podcast scripts, hooks, and CTAs for short-form (TikTok, Reels, Shorts), YouTube long-form, or podcast episodes. Triggers: "write me a reel about X", "I need a hook for my video", "script a 60-second video", "draft a podcast episode", "turn this idea into a script". For motion/timing/shot specs rather than the words, use motion-picture.
One small anthropomorphic character carries the whole film. Spell out its looks part by part, write the style as render settings you can measure, and cut the timeline so each block holds a single action goal. Use when the user wants a Seedance 2.5 / 2.0 video prompt for this kind of clip (3D cartoon character short; Chinese trigger phrases: 3D 卡通角色短片, 3D 卡通, 皮克斯, 黏土, 卡通角色), or asks how published cases of this type were prompted. Read references/cases.md before drafting.
Use when the user asks for text-to-speech narration or voiceover, accessibility reads, audio prompts, or batch speech generation. OpenAI remains the default; Atlas Cloud is an explicit optional backend for asynchronous multilingual speech. Custom voice creation is out of scope.
Download TikTok videos browser-natively — no `yt-dlp`, no signature solver, no watermark. The browser plays the video (URL signing, CDN tokens, quality selection all done by the page itself); ttdl records the demuxed media the player feeds to MediaSource via a `SourceBuffer.appendBuffer` hook and muxes to MP4 with ffmpeg. If the user can watch the video, it's downloadable — region-locked and age-gated content work as long as a logged-in tab can play them. The capture is the clean, unwatermarked playback stream (TikTok's own Download button serves a separately-rendered watermarked file). Use when the user wants to download, save, or fetch a TikTok video (with or without audio) to disk.
Controls the Unity Editor and projects from the command line, driving a running Editor to edit scenes and assets or run C#. Use when asked to install or manage Editors, licenses, or projects, set up version control, build, test, configure the Unity MCP server, or run any unity command.
Use this skill for AI video generation. Triggers include: "generate video", "create video", "make video", "animate", "text to video", "video from image", "video of", "animate image", "bring to life", "make it move", "add motion", "video with audio", "video with dialogue" Supports text-to-video, image-to-video, video with dialogue/audio using Google Veo 3.1 (default) or OpenAI Sora.
Create high-fidelity web demo animations for video, course and narration scripts, with one HTML per shot (storyboard demo pages, video companion demo animations, animated web pages for screen recording, visualised shots for science, tech and product launches). Use this Skill whenever the user mentions making a demo animation web page for a video, storyboard HTML, shot pages, turning a voiceover script or outline into animation, designing video visuals as web pages, recording a web animation, or reworking and upgrading an existing batch of demo pages. The core is a unified cinematic player mechanism + a completely different visual style per page + shot language + synchronised sound effects + character expression commentary, with content fully open and unrestricted.
Video processing toolkit. Use when user wants to: - Download videos from YouTube or other sites - Remove silence from videos - Trim, cut, or extract segments from videos - Extract audio from video files - Enhance or denoise audio - Replace audio track in a video - Change video playback speed - Concatenate multiple videos - Generate transcripts/captions (VTT) - Generate video descriptions, timestamps, or context cards - Upload videos to YouTube or Bunny.net CDN - Post social updates to X (Twitter) or LinkedIn - Get video metadata (duration, resolution, codec)
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.