Skills
97,222 packaged capabilities and playbooks for agents. Install any of them with one command.
Creative & media
Install and wire registry blocks and components into FrameVideo compositions. Use when running framevideo add, installing a block or component, wiring an installed item into index.html, or working with framevideo.json. Covers the add command, install locations, block sub-composition wiring, component snippet merging, and registry discovery.
Core graphic design skill for one-off visuals via Gemini 3.1 Flash image generation and/or HTML/CSS/SVG code. Covers composition, color theory, typography, prompt crafting, exact-size cropping, output verification, text-overlay fallback, and advanced CSS techniques. Use for general or custom-size graphics that don't fit a specific format; for platform formats prefer the sibling skill (social-media-graphic, poster-design, thumbnail-design, ad-creative-design, banner-design, product-mockup, infographic). Trigger phrases: "design a graphic", "create a visual", "make a graphic", "design an image", "create artwork", "generate a graphic", "visual design".
GSAP animation reference for FrameVideo. Covers gsap.to(), from(), fromTo(), easing, stagger, defaults, timelines (gsap.timeline(), position parameter, labels, nesting, playback), and performance (transforms, will-change, quickTo). Use when writing GSAP animations in FrameVideo compositions.
Turn a user-provided voiceover audio file + SRT subtitle into a finished, narration-synced MP4 using HyperFrames (HTML-to-video). The audio + SRT are the source of truth — scenes are timed to the SRT cues, content is read from the SRT, and the audio is muxed in automatically. Use when the user hands over an mp3/wav + srt and wants a video, says "把音频做成视频", "做一期视频", "audio + srt to video", "把这期早读做成视频", "render this narration into a video", or provides a recording + subtitles for an explainer / daily / 解读 / 口播. NOT for generating the voiceover (that's the user's job here) and NOT for slide decks (use a slides skill).
Polishes speech transcripts (interviews, speeches, podcasts, meetings) into more readable article paragraphs. Use when the user mentions “字幕精修”, “transcript polish”, “润色字幕”, “把视频字幕整理成文章”, “访谈文字整理”, processing interview notes, optimizing transcripts, organizing speech-to-text output, or turning long dialogue or speech transcripts into readable articles. Suitable for transcripts of solo speeches or multi-person conversations; preserves the original sentences and wording and avoids excessive summarization. Use this skill even if the user simply says “帮我整理一下这段文字” and provides clearly colloquial text.
Pull original ("vanilla") Minecraft textures from Mojang's bedrock-samples repository through the jsDelivr CDN and apply them to Blockbench models, resolving block IDs to per-face files, tinting grayscale grass/leaves/water, tiling large faces, handling flipbooks, and importing the vanilla PBR texture_set (MERS/normal). Use when the user asks for vanilla, default, original, real or Minecraft-accurate textures, names a Minecraft block or material for a model (brick wall, oak log, stone bricks, grass block, wool, glass), wants a model to match the base game, or needs a Bedrock texture name mapped to Java Edition. Works for Bedrock, Java and generic models.
Create an AI persona from scratch. Interactive flow: enter a name → provide source material (URL/file/pasted text) → automatic distillation → output a ready-to-use skill directory. Activate when the user says /create-soul, “创建人物,” “蒸馏 [某人],” or “create a persona for [name]”
End-to-end video production orchestration for the content-creation workspace. Use when the user asks to make, recreate, package, render, quality-check, or archive a video; says 制作待制作队列, 按交接稿制作, or 制作队列视频; points to a handoff project folder under 01-内容生产/视频工作台/待制作/ (with contract file 交接稿.md inside) or an image-card handoff under 01-内容生产/图文工作台/待制作/ (whose frontmatter is the binding production contract); when a request could involve HyperFrames, Remotion, captions, TTS, downloading/transcription, website/product videos, motion graphics, or reference replication; or when the user needs production planning before editing.
A general-purpose self-media content production and operations workflow. Use for content creation, topic planning, trending-topic or competitor research, adapting content for multiple platforms, short-video scripts, digital-human videos, animated explainers, publishing videos across YouTube, Douyin, WeChat Channels, and Xiaohongshu, publishing packages, WeChat Official Account drafts, content analytics, weekly and monthly reviews, and continuing unfinished tasks. Identifies request types, invokes modules for creative briefs, content strategy, trends and competitors, platform copy, short videos and video production, video publishing, analytics reviews, and delivery archiving, while managing direction, title, final-copy approvals, and publishing authorization.
Render, preview, validate, and burn consistent production subtitles from QC-passed real-timestamp caption artifacts. Use for AI-video production, washed-video production, IndexTTS2 explainers, practical long-form videos, talking-head edits, Kimi-style dark rounded captions, warm-white translucent captions, subtitle styling, subtitle overlays, or fixing caption size, position, shadow, drift, and 4K scaling. Always pair with ra-audio-to-subtitles when final word timestamps do not already exist.
Tailwind CSS v4.2 browser-runtime patterns for FrameVideo compositions. Use when scaffolding or editing projects created with `framevideo init --tailwind`, writing Tailwind utility classes in composition HTML, adding CSS-first Tailwind v4 theme tokens, debugging v3 vs v4 syntax, or deciding when to compile Tailwind to CSS instead of using the browser runtime
TypeGPU and raw WebGPU adapter patterns for FrameVideo. Use when creating GPU-rendered compositions with TypeGPU, raw WebGPU, WGSL fragment shaders, compute pipelines, liquid glass effects, particle systems, or any canvas layer driven by navigator.gpu that responds to FrameVideo fv-seek events
Web Animations API adapter patterns for FrameVideo. Use when authoring element.animate() motion, Animation currentTime seeking, document.getAnimations(), KeyframeEffect timing, fill modes, or native browser animations that must render deterministically in FrameVideo
Translate reference images or brief descriptions into editable halftone paper-collage specifications. By default, use Codex built-in image_gen to create brand-consistent stills and transparent layers, then use local FFmpeg or HyperFrames to make deterministic motion videos that assemble piece by piece from an empty background. Use for requests involving halftone collage, paper collage, paper-cut collage, editorial collage, magazine collage, vintage print collage, collage motion, collage animation, paper-cut animation, assembling paper pieces, layer-by-layer assembly, building from nothing, elements flying in to assemble, stop-motion collage, stop-motion animation, Arcads style, breaking down a reference image or style, reverse-engineering a visual look, making something in this style or something similar, generating collage stills, animating an image or assembling elements, or creating collage ads or explainer videos. English trigger phrases include editorial collage, halftone collage, paper collage, paper-cut motion, stop-motion graphic, assemble from empty, build-on motion, analyze this, break this down, reverse-engineer this look, generate the stills, animate this, and make these move. Do not call Arcads, Nano Banana, Seedance, or Kling by default; use an external generation path only when explicitly requested and the corresponding connector is available
Produce a polished local talking-head or narrated screen-recording rough cut without a cloud editor. Use when Codex must clean Chinese or mixed Chinese-English speech, correct product terminology before semantic editing, compress pauses without making speech breathless, preserve source resolution and frame rate, normalize dialogue loudness, generate final-audio subtitle artifacts, or benchmark local output against ChatCut/video-use/chengfeng/AI剪口播.
Generate professional infographics with 21 layout types and 20 visual styles. Analyze content, recommend layout and style combinations, and create publishable infographics. Use when the user asks for an “infographic”, “visualization”, “high-density information graphic”, or “visual summary”
Video Pre-Launch Check Officer. Used for requirements preparation, asset inventory, design specification confirmation, storyboard breakdown, and pre-render review before video generation. Use when the user wants to make a video, promo, product demo, animated short, Douyin/YouTube content, or revise storyboards, pacing, shots, subtitles, voice-over, or transitions. Outputs a standardized video-spec.md and generates a preview-board.md table of keyframe previews for user approval before HyperFrames generates the video, reducing rework caused by direct rendering.
An Agent skill that turns long Markdown articles into 3:4 (1080×1440) image sets for Xiaohongshu. Includes five design skins (E Elegant Journal · default / A Warm Apricot Latte / B Misty Pine Green / C Ink and Gold / D Gilded Night Reading) plus three women-oriented themes (F Pink Kingdom / G Otome Game Night / H Ghibli) × multiple page layouts. The Agent reads the article, splits it into pages, and fills in HTML; the Playwright renderer captures and exports each image as PNG. Use when the user says “turn a long article into images”, “Xiaohongshu long image”, “turn this article into Xiaohongshu graphics/images/cards”, “article to images”, “typeset with my design draft”, “convert to a 3:4 image set”, wants a “women-oriented/girly/pink/otome-game/Ghibli-style” layout, or provides an article and mentions posting it on Xiaohongshu.
Process and generate multimedia content using Google Gemini API for better vision capabilities. Capabilities include analyze audio files (transcription with timestamps, summarization, speech understanding, music/sound analysis up to 9.5 hours), understand images (better image analysis than Claude models, captioning, reasoning, object detection, design extraction, OCR, visual Q&A, segmentation, handle multiple images), process videos (scene detection, Q&A, temporal analysis, YouTube URLs, up to 6 hours), extract from documents (PDF tables, forms, charts, diagrams, multi-page), generate images (text-to-image with Imagen 4, editing, composition, refinement), generate videos (text-to-video with Veo 3, 8-second clips with native audio). Use when working with audio/video files, analyzing images or screenshots (instead of default vision capabilities of Claude, only fallback to Claude's vision capabilities if needed), processing PDF documents, extracting structured data from media, creating images/videos from text prompts, or implementing multimodal AI features. Supports Gemini 3/2.5, Imagen 4, and Veo 3 models with context windows up to 2M tokens.
Builds cinematic motion graphics in the browser (promo videos, openers, intros, bumpers, kinetic typography, 16:9 or 9:16 explainer videos) using HTML + CSS + GSAP (plus Three.js when needed), with results that move like a real video — not a presentation slide deck. Use this skill whenever the user asks for a "promo video", "opener", "animated intro", "motion graphic", "bumper", "kinetic typography", "cinematic text animation", "explainer", or "explainer video" (visual journalism, educational cartoon, collage), or shows a promo/explainer video reference and wants a web version made — even if they don't mention the words "motion graphic" explicitly. Also use it when the user complains that a web animation looks "like a PPT/presentation" and wants it more cinematic, wants to render a web animation to an MP4 video file, or wants a cartoon/animated explainer built DIRECTLY in After Effects via the bridge/MCP Higgsfield.
An AI image-processing and e-commerce product image generation skill pack (美图设计室 DesignKit). Supports background removal, transparent backgrounds, AI sharpening and image restoration, and generation of product hero images and Listing image sets; routes based on user intent to designkit-edit-tools and designkit-ecommerce-product-kit.
Expert real-time VFX artist specializing in particle systems, shader effects, and the invisible craft that makes games feel satisfying. Masters Niagara, VFX Graph, Godot GPU particles, and understands the AAA principles that make effects read clearly at 60fps. Use when "particle system, visual effects, vfx, particles, niagara, vfx graph, flipbook, sprite sheet, explosion effect, magic effect, trail effect, beam effect, dissolve, distortion, force field, hit effect, muzzle flash, impact effect, smoke particles, fire effect, soft particles, game juice, screen shake, particle overdraw, effect optimization, vfx, particles, effects, niagara, vfx-graph, game-juice, visual-effects, shaders, flipbook, trails, beams, explosions, optimization, gpu-particles" mentioned.
Use when the user has a long-form video (interview, lecture, podcast, or conversation) and an SRT transcript, and wants to extract 3–6 stand-alone topical short clips. This skill ONLY cuts and crops—it produces raw clips plus per-clip SRTs as a hand-off package for downstream post-production (`/wjs-overlaying-video`). Triggers: “切成几段”, “分主题”, “拆成短视频”, “切片”, “topic segments”, “split into clips”.
Deal Flow editorial skill — signal composition, source validation, and editorial voice guide for aibtc.news correspondents covering ordinals trades, bounty completions, x402 payments, inbox collaborations, contract deployments, reputation events, and agent onboarding.
Create subtitles for edited talking-head footage: calculate post-edit timing directly from the existing word-by-word transcript and ledger (no export or re-transcription needed), correct misheard proper nouns using the dictionary and author’s manuscript, split by sentence, and review each screen in Studio. Use when the user asks to create, add, or edit subtitles, re-split them, or says the subtitles are wrong. Do not use for deleting words, physical cuts, storyboard animation, or final rendering.
Automate and script DaVinci Resolve workflows. Use when a user asks to script DaVinci Resolve via Python/Lua API, automate color grading, batch render projects, manage timelines programmatically, automate media import, build render queues, create Fusion compositions via script, automate Fairlight audio processing, manage project databases, build custom tool scripts, or integrate Resolve into production pipelines. Covers the Resolve Scripting API (Python/Lua), Fusion scripting, and workflow automation.
Edit one chapter or section of an already-generated video without redoing the whole thing. Change the script, visuals, captions, or timing of a single scene, then re-render and re-verify - iterate on AI-generated videos, social clips, promos and explainers. Companion to make-video. Use when asked to change, fix, redo or tweak a section, scene or chapter of an existing video. Trigger - /edit-video <project-id> <chapter-id> "<instruction>"
Create and composite HeyGen digital-human videos in the content workspace. Use whenever the user asks for a 数字人, 数字人5, HeyGen, AI 口播人物, 数字人小窗, 圆形头像口播, full-screen digital human, or asks to make a video that uses the existing avatar. Enforce Creator-plan CLI OAuth, the configured Digital Twin, local IndexTTS2 narration, Avatar III for an unspecified digital human but Avatar V whenever the named preset 数字人5 is requested, adaptive face measurement plus the approved human-reviewed visual-centering offset in the headroom_08 circular bottom-left layout unless full-screen is explicitly requested, subtitle-safe placement, credit-aware master reuse, final QC/archive routing, and mandatory pairing with `heygen-avatar-motion` before any motion prompt or paid generation.
Create programmatic vector animations with TypeScript using the Motion Canvas library. Use when building animated explainer videos, tutorials, data visualizations, or presentations with synchronized audio. Triggers on tasks involving motion graphics, animation scenes, code-based video creation, or requests to animate shapes/text/images with precise timing control.
Podcast-to-everything content pipeline. Takes a podcast RSS feed or raw transcript and generates a full cross-platform content calendar: short-form video clips, Twitter/X threads, LinkedIn articles, newsletter sections, quote cards, blog outlines with SEO keywords, and YouTube Shorts/TikTok scripts. Scores each piece by viral potential (novelty × controversy × utility) and deduplicates against recent output. Use when asked to: “repurpose this podcast,” “turn this episode into content,” “podcast content calendar,” “extract clips from this episode,” “podcast to social,” “content from RSS feed,” “batch process episodes,” or any request to turn podcast/audio content into a multi-platform content plan
Use when creating or revising finished video thumbnails or cover images with Scenario: game trailers, boss reveals, patch updates, devlogs, or community videos needing readable composition, exact headline text and consistent characters. Not for title ideas or critique alone, in-video title cards, or a generic image without a thumbnail deliverable
An integrated tool for generating professional Seedance 2.0 storyboard prompts and creating videos. Use when users want to create storyboard videos, generate videos with Seedance/即梦, need professional storyboard prompts and direct video generation. Supports multiple image references, storyboard guidance, API-based video generation, and automatic downloads.
Generates professional infographics with various layout types and visual styles. Analyzes content, recommends layout and style, and generates publication-ready infographics. Use when user asks to create "infographic", "信息图", "visual summary", or "可视化".
Generates professional infographics with 21 layout types and 20 visual styles. Analyzes content, recommends layout×style combinations, and generates publication-ready infographics. Use when the user asks to create "infographic", "信息图", "visual summary", "可视化", or "高密度信息大图".
Writing plays and screenplays in the Chekhovian mode (契诃夫戏剧法) — distilled from the complete plays (没有父亲的人/普拉东诺夫, 伊凡诺夫, 林妖→万尼亚舅舅, 海鸥, 三姊妹, 樱桃园, nine one-acts) plus Chekhov's own letters, Stanislavsky's and Nemirovich-Danchenko's rehearsal notes and 焦菊隐/童道明 criticism. Covers the four-act mood-based structure without central climax, off-stage events, "correct answer stated in Act 1 and refused", the announced event that never happens, three-layer endings (hope monologue + nihilist line + household action), no-villain conflict with environment, every character's object and catchphrase, one-way love chains, half-step-off dialogue, philosophy cut by thirst, the shortest words for the deepest pain, stage-direction and sound discipline, and how Chekhov revised The Wood Demon into Uncle Vanya. Use when the user wants an ensemble drama, a "nothing happens" play, subtext-heavy dialogue, a family estate / provincial story, an anti-climactic structure, or asks how to "Chekhov-ize" a script.
Senior creative director skill: generates concepts with SIT, TRIZ, SCAMPER, bisociation and lateral-thinking methods, scores them against weighted criteria calibrated on Cannes Lions, D&AD and HumanKind standards, and refines until the bar is met. Takes briefs as text, notes or PDFs.
AI-native video editor for any agent. Use when the user asks to edit, trim, splice, caption, zoom into, add a title card, transition between, remove silence from, chroma-key, stabilize, or render video files. Wraps FFmpeg in 19 deterministic tools and a session log; runs locally; no generative AI.
A provider-independent FFmpeg finishing workflow for AI agents preparing generated or edited media deliverables. Use when finalizing images, image sequences, video, audio, captions, overlays, social/export variants, checksums, manifests, delivery specs, and QA with ffmpeg/ffprobe. Covers transcode versus stream-copy decisions, scaling, frame-rate handling, color metadata cautions, loudness normalization, subtitle burn-in or sidecars, concat/trim, GIFs, batch variants, and reproducible command reporting.
Use when writing or fixing Godot 4.7 shaders: .gdshader code (spatial, canvas_item, particles, sky, fog), VisualShader graphs built in code, stylized or toon water, dissolve or burn effects, hit flash, outlines, X-ray or portals with the stencil buffer, depth or screen-texture post-processing, global or per-instance uniforms, compute shaders and Texture2DRD, the shader baker; or when 'shader compiles but looks wrong', 'black screen from my post effect', 'my Godot 3 shader breaks', 'colors look washed out', 'first-launch shader stutter'.
Use when making or fixing particle effects and game feel in Godot 4.7: GPUParticles3D or GPUParticles2D explosion, fireball projectile with trail, impact, sparks, smoke, shockwave, ribbon trails, sub-emitters, particle collision, color curve does nothing, particles pop out or vanish, effect too heavy on mobile, CPUParticles fallback for Compatibility or web, overdraw, hit-stop, freeze frame, screen shake, juice.
Edit any video by conversation. Transcribe, cut, color grade, generate overlay animations, burn subtitles — for talking heads, montages, tutorials, travel, interviews. No presets, no menus. Ask questions, confirm the plan, execute, iterate, persist. Production-correctness rules are hard; everything else is artistic freedom.
Score and compare images using vision LLMs as judges. YAML-defined criteria presets for 11 use cases (text-to-image, photorealism, document OCR, charts, UI, portrait, product, scientific, invoice, alt-text, artistic style). Supports OpenAI, Anthropic, Gemini, Mistral, and OpenRouter as judge providers. Keys auto-decrypted via SOPS + age.
Transforms content (URLs, uploaded documents, pasted text, meeting transcripts) into professional visualizations across four output modes. Accepts a mode argument or a keyword trigger in the user message. Mode "diagram" produces an Excalidraw diagram via Excalidraw:create_view. Mode "infographic" generates a Swiss Pulse PNG via the Gemini image-generation API. Mode "visualize" renders an inline Visualizer widget (SVG or HTML) via visualize:show_widget. Mode "publish" ships an interactive Swiss Pulse HTML visual to HeyGenverse via HeyGenverse:create_app and returns a shareable link. Keywords that activate the skill: "diagram it", "excalidraw this", "draw a diagram of this", "nano this", "vis it", "ver it", "hey it", "heygenverse this". Do not use for plain-text summaries, code explanations, prose responses, or generic chat visualizations without a chosen output format.
Generates images and text via reverse-engineered Gemini Web API. Supports text generation, image generation from prompts, reference images for vision input, and multi-turn conversations. Use when other skills need image generation backend, or when user requests "generate image with Gemini", "Gemini text generation", or needs vision-capable AI generation.
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.