Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Creative & media
Generate AI images using Gemini image generation API. Use this skill when content needs images - thumbnails, social posts, blog headers, or creative visuals. Follows an iterative workflow - brainstorm concepts, select direction, generate in multiple styles, then produce via API
Use this skill whenever the user asks for help with music composition, music theory, songwriting, harmony, melody, counterpoint, arrangement, orchestration, or musical analysis — across classical, jazz, pop, rock, hip-hop, R&B, electronic, film/TV, K-pop/J-pop, folk, or musical theatre. Triggers include explicit theory questions (“what scale fits over Cm7”, “explain ii-V-I”, “modulate from C to E♭”), craft questions (“write a chord progression for a melancholy bridge”, “make this melody more memorable”, “voice this chord on piano”), and vague creative problems that map to compositional techniques (“the chorus feels weak”, “this transition is awkward”, “my song sounds generic”). Use this skill even when the user describes the goal in non-technical terms — translating fuzzy creative problems into concrete compositional moves is exactly what this skill is for. Do NOT use for DAW operation, MIDI file generation, audio engineering, mixing/mastering, sound design, or notation software UI.
Turn a novel into a five-part short drama outline package: an adaptation brief, character list, highlight moments list, episode-by-episode summaries, and an asset inventory. Produce outline.json + Markdown + a one-page review report (KPI band, key decisions, highlight timeline, scheduling matrix, scene overview, quality gates). All 14 quality gates are checked deterministically by scripts (character tier limits, main-scene limits that vary dynamically with episode count, no more than 3 episodes between highlight moments, required suspense hooks for every episode, etc.); they don’t rely on the model to remember. Supports a checkup mode: provide an existing outline to run only the quality gates and get a diagnosis. Zero dependencies, no API key; uses the current session quota. Use when asked for an adaptation outline, short drama outline, outline breakdown, novel-to-short-drama adaptation, or adaptation outline.
Resize photos and videos to exact pixel dimensions or aspect ratios using Adobe tools. Use this skill whenever a user wants to resize, scale, or change the dimensions of an image or video file — including phrases like "resize this to 1920x1080", "make this 4K", "scale to 800x600", "change the aspect ratio to 16:9", "resize my video", "make the image smaller", "crop to square", "fit this to a specific size", "resize for print", "resize for web", "make it 300 DPI ready", "change canvas size", "resize a batch of photos", or any request specifying target dimensions (W×H, ratio, or named size like "4K", "HD", "A4"). Also triggers for: "make this fit a specific size", "resize to [any dimension]", "I need this at [WxH]", "scale my video down", "change resolution", "downscale", "upscale". NOT for social media platform sets (use adobe-create-social-variations for that). Uses image_crop_and_resize for photos, video_resize for videos
Captions living in 3D space around a talking head: a camera flies between the speaker and the words (whip-in, parallax truck, push), caption groups sit at different depths so camera moves pull them apart, hero words hide BEHIND the speaker through an alpha matte, a ring of words wraps round the speaker and turns in front of them, focus racks between depths, and text steps at 15 fps with ghost motion blur. Works in any type style (editorial serif, bold sans) or hand-drawn (p5.brush write-on). Trigger on: "3D text", "3D camera captions", "text around / behind the speaker", "camera moves through the text", "depth captions on my avatar video". Covers prep (portion, padded plate, matte, word clock, font metrics), the shot grammar, depth rules, QC, and limits.
Plan and build 2D game maps and scenes - top-down and side-scrolling levels, tilemaps, prop packs, parallax backgrounds and HD-2D battle or story plates - with collision, navigation and reachability checks, environment motion, a playable HTML preview and export to Tiled, Godot 4 or LDtk. Map art (terrain, tiles, props, plates, parallax layers) is image generation by default, through generate2dmedia route_media.py (the API when a key is configured, else the local Codex or Grok CLI); map data (layout, collision, exits, spawns) is written as data. Use for any map, level, room, scene, background, terrain or prop-kit request. Not for characters, creatures, items or FX (generate2dsprite), character animation (video2dsprite), code-drawn tiles or maps (codeart2d, only on request) or calling an image or video API by itself (generate2dmedia)
Create animated financial candlestick charts in Remotion projects. Covers data retrieval (Yahoo Finance), plugin architecture (config / timeline / Scene / plugin), frame-by-frame Canvas drawing, a dark trading-terminal style, and Remotion Studio preview and render output. Triggers: candlestick charts, candlestick, OHLC, stock animations, financial chart videos, Remotion charts.
Web Audio framework for creating interactive music in browsers. Use when building audio applications, synthesizers, musical instruments, effects processors, audio visualizations, DAWs, step sequencers, or any browser-based sound generation. Handles synthesis, scheduling, sample playback, effects, and audio routing.
Upload videos to Chinese and international social media platforms using browser automation. Supports Douyin (抖音), Kuaishou (快手), Xiaohongshu (小红书), TikTok, and Tencent Video/WeChat Channels (视频号). Use when users need to automate video publishing to these platforms with complete field generation and anti-detection browser configuration.
Route short-form video prompt work into the right hook or segment pattern before writing storyboards or provider requests. Use when the user has a brief, a segment type such as hook/benefit/cta/creator, duration constraints, and optional references, and you need to choose the strongest narrative structure instead of freehand prompting.
Resize, crop, or export any image or video into platform-ready social media assets using Adobe Creative Cloud tools. Use this skill when a user wants to prepare a photo, image, or video for one or more social platforms—Instagram, TikTok, LinkedIn, Facebook, YouTube, Snapchat, Pinterest, Threads, or X/Twitter. Triggers on: “prepare my image for Instagram”, “resize for TikTok”, “get this ready to post”, “make versions for all platforms”, “social media sizes”, “crop for stories”, “export for LinkedIn”, “resize my video for social”, “make social media assets”, or any request to adapt a photo or video for specific platforms. Handles subject-aware cropping, AI canvas expansion, test previews before full runs, and same-ratio video resizing.
When output contains Chinese characters and punctuation, when writing Chinese documents, generating Chinese interactive responses from Chinese documents, replying to user questions or tasks in Chinese, or working with mixed Chinese and non-Chinese text and punctuation, this Skill must be used to receive professional guidance on Chinese character usage and behavioral standards
Provider-independent color grading and finishing direction for generated video, ads, trailers, product films, social clips, explainers, and mixed-source edits. Use when planning, directing, reviewing, or QAing color correction, shot matching, exposure, contrast, saturation, skin/product color protection, look development, LUT/reference use, SDR/HDR color management, delivery constraints, generated-media artifacts, editor/compositor handoff, and finishing QC
Analyzes events through a narrative lens using story structure, character arc analysis, dramatic tension, thematic development, and narrative theory (three-act structure, hero's journey, conflict-resolution). Provides insights on narrative coherence, character motivations, dramatic stakes, plot development, and thematic resonance. Use for complex human stories, leadership analysis, organizational narratives, crisis narratives, and cultural moments. Evaluates character development, narrative arc, dramatic tension, thematic depth, symbolic meaning, and narrative coherence.
Design and critique TikTok/Reels visual hooks for short-form slideshow or video production. Use this when discussing how to make the first image, first frame, opening shot, visual sequence, slideshow cover, or first 1-3 seconds strong enough to stop the scroll, especially for generated image/video workflows and UGC-style ads
Create storyboards for AI short dramas: a three-level structure—segment (one video generation, ≤15 seconds) → shot (2–5-second cuts within a segment, each assigned a screenplay beat) → storyboard frame (one keyframe per cut: the main frame is pinned at 0.00 seconds, and child frames are pinned at their respective cut points). Each segment comes with a MiniMax H3 video prompt (official guidance defaults to English, one line per shot; promptLang can switch to Chinese): alignment instructions and [Shot k] cut timestamps are derived from the storyboard structure and checked character by character; dialogue is included verbatim in <d> blocks (conventions are embedded in references/h3-prompt.md, with no external skill dependency). Outputs storyboard.json + Markdown + a single-page review report (shot pacing track / episode shot list / generation batch sheet / voiceover alignment sheet, including JSON export). Does not generate images—the deliverables are each frame’s visual prompt and the reference images it should link to. All 18 quality gates are checked deterministically by scripts (gate 18, shot-recipe, is optional: it is checked only if the shot-recipes card library is attached; otherwise the skip is stated explicitly); one-click export creates an H3 production package (prompts per segment + storyboard image list ordered by Picture). Zero dependencies, zero API keys; uses the current session quota. Use when asked to create a storyboard, shot list, or cuts for an AI short drama
Generate fashion lookbooks, model showcases, and style-video prompts for Seedance 2.0 (Higgsfield). Use when the user wants fashion video content, lookbook videos, model runway walks, outfit showcases, style guides, fashion events, runway clips, street-style content, or any fashion or clothing video. Trigger for fashion videos, lookbooks, model showcases, clothing, style guides, fashion events, runways, street style, collection launches, fashion ads, outfit videos, OOTD, fashion films, or any fashion or model video request. This also applies to requests such as “showcase this outfit” or “create fashion content for my brand.”
Use this skill when the user asks to "create a style sheet", "style guide", "house style", "keep the voice consistent", "voice drift", "British or American spelling", "character voices", "lint the prose", "prose check", "filter words", "said-bookisms", "overused words", "repeated phrases", "similar character names", "voice fingerprints", or wants to record and enforce the voice and surface conventions of a story project. NOT for rewriting dialogue to make the voices distinct when everyone sounds the same (use line-editing), or a character's profile or arc (use character-management).
Run AI apps via inference.sh CLI - image generation, video creation, LLMs, search, 3D, Twitter automation. Models: FLUX, Veo, Gemini, Grok, Claude, Seedance, OmniHuman, Tavily, Exa, OpenRouter, and many more. Use when running AI apps, generating images/videos, calling LLMs, web search, or automating Twitter. Triggers: inference.sh, infsh, ai model, run ai, serverless ai, ai api, flux, veo, claude api, image generation, video generation, openrouter, tavily, exa search, twitter api, grok
Transcribe non-realtime speech with Alibaba Cloud Model Studio Qwen ASR models (`qwen3-asr-flash`, `qwen-audio-asr`, `qwen3-asr-flash-filetrans`). Use when converting recorded audio files to text, generating transcripts with timestamps, or documenting DashScope/OpenAI-compatible ASR request and response fields.
Volcengine AI MediaKit audio and video processing skill. Triggered when users need to process or edit audio/video content. After processing, it automatically checks task status and returns playback links for generated outputs. Core capabilities fall into five categories: 1) Video processing: multi-clip stitching, clip trimming, frame flipping, video speed adjustment, audio speed adjustment, image-to-video generation, audio-video composition, audio track extraction, and audio mixing; 2) Audio processing: vocal/accompaniment separation and audio noise reduction; 3) Video enhancement: comprehensive quality restoration, AI super-resolution, and intelligent frame interpolation; 4) AI content analysis: ASR speech-to-text, OCR text extraction, subtitle removal, subtitle embedding, intelligent scene slicing, portrait matting, green screen matting, media info query, and highlight extraction; 5) AI content generation: comic style transfer, AI video translation, AI drama recap narration, and AI drama script restoration. Trigger keywords: video stitching, video trimming, video editing, video speed change, video flipping, image to video, audio-video composition, audio extraction, mixing, vocal separation, background sound separation, audio denoise, denoise, AI super-resolution, super resolution, quality restoration, quality enhancement, intelligent frame interpolation, video frame interpolation, frame rate increase, speech recognition, ASR, OCR, text extraction, subtitle removal, subtitle erasure, add subtitle, embed subtitle, scene segmentation, scene slicing, portrait matting, green screen, highlight, comic style, video translation, drama recap, drama narration, drama script, script restoration, media info. Not applicable to pure text generation, real-time streaming processing, video understanding/analysis (except supported AI features), or AI generative video creation without source media inputs.
Generate elegant cover images for articles. Analyzes content and creates eye-catching hand-drawn style cover images with multiple style and composition options. Use when the user asks to “generate cover image,” “create article cover,” or “make a cover for article”
Build fact-grounded short-form video personas and visual consistency packs from validated benchmark research. Use when defining a repeatable creator archetype, image prompt pack, or persona lock for batch video production. Personas must be derived from real benchmark evidence, such as creator types, protagonist descriptions, visual styles, hooks, and audience language. Do not invent personas or visual traits without source support.
Write scripts for AI short dramas: turn novel-outline episode summaries into structured scenes and beat sequences, alternating action beats and dialogue lines, with every line tagged by speaker and tone. Episode durations are calculated deterministically from speaking speed. Produces script.json, Markdown, and a one-page review report (duration dashboard, episode scripts, scene overview, and dialogue grouped by character, with JSON export). The script manages the drama; storyboards manage the shots—this skill does not create storyboards, shot numbers, or generation prompts; those belong to the next-layer storyboard skill. All 10 quality gates are checked deterministically by scripts (episode duration ±15%, single lines ≤35 characters, valid speakers, hooks written down, concrete hook payoff within the first 3 beats (hookBeat claims the cold open), at least one action beat per scene, payoff claims, and reconciliation of characters/scenes/lighting/props with upstream data…). If outline.json exists, use seed to prefill each episode's skeleton; with --art, reconcile lighting states too. Zero dependencies, zero API keys; uses the current session quota. Use when asked to write a script, dialogue, scenes, or screenwriting for an AI short drama.
Produce slides-driven narration videos (口播视频) where each slide maps 1:1 to one voiceover section. Orchestrates a slides-generating skill (PPT, chosen from whatever is available) and `video-planner` (script + publishing materials) with a method-focused production workflow. Use when user wants to make a video that uses slides to explain a topic - e.g. 发布解读 / 产品评测 / 行业观察 / 技术解读 / 趋势分析. Triggers on "做一期视频 + PPT", "slides 视频", "发布解读视频", "深度讲解视频", or similar requests for structured narration videos.
Add 3D models (.glb/.gltf) to a Decentraland scene using GltfContainer — 8,800+ free assets from the OpenDCL catalog and Creator Hub asset packs (Genesis City, Halloween, Sci-fi, Fantasy, Smart Items; the latest additions are catalogued offline), plus authoring and editing custom models by driving Blender (headless CLI or Blender MCP). Use when the user wants to add models, import GLB files, find free 3D assets, set up model colliders, or create/edit/retexture/optimize a model in Blender — also whenever an `mcp__blender__*` tool is available and the task involves 3D models. Do NOT use for SDK materials/textures (see advanced-rendering), model animations (see animations-tweens), or avatar emotes (see avatar-animations).
Analyzes video files by extracting frames, detecting scene changes, and optionally transcribing audio. Produces structured markdown analysis with timeline, visual descriptions, and key moments. Use when the user provides a video file (.mp4, .mov, .avi, .mkv, .webm) and wants analysis, asks to "analyze a video", "describe this video", "video timeline", "what happens in this video", or mentions video content understanding. Supports custom frame rates ("at 5fps", "1 frame every 3 seconds"), time ranges ("from 0:10 to 0:30", "first 15 seconds"), visual-only mode ("skip audio"), and user-provided transcripts.
Match spoken edit beats to candidate B-roll assets using a normalized transcript, subtitle chunking, optional A-roll analysis, and a reusable B-roll catalog. Use this when the goal is to decide what B-roll should support each beat, not just to list assets or describe the video
Expert game audio designer and implementer specializing in interactive sound design, adaptive music systems, spatial audio, and audio middleware integration. Brings deep knowledge of FMOD, Wwise, and native engine audio systems to create immersive sonic experiences that respond dynamically to gameplay. Use when "game audio, sound design, game music, FMOD, Wwise, spatial audio, 3D sound, audio middleware, game sfx, adaptive music, interactive audio, audio bus, game mixing, audio occlusion, reverb zones, audio pooling, sound manager, audio, sound, music, game-audio, fmod, wwise, spatial-audio, middleware, mixing, sound-design" mentioned.
Use Seedance 2.0 (Higgsfield) to turn comic panels, comic pages, webcomics, and illustrated storyboards into animated videos. Use when users want to animate comics, bring illustrations to life, convert comics to video, animate storyboards, or create motion from static sequential art. Triggers: comic to video, comic animation, panel animation, storyboard to video, webcomic animation, comic book motion, sequential art, graphic novel animation, or any illustration-to-animation request. Also use when users say “make this picture move” or “animate this page.”
Distill reusable Writing DNA from at least 20 complete articles for Chinese or English authors, publications, brands, and accounts; use for language, structure, topic logic, source strategy, cognitive-frame, and visual-style analysis.
Turn product information, packaging and qualifications, target platform, audience, and visual references into actionable e-commerce visual plans, including conversion strategy, Campaign Style Lock, main images/detail pages/Listing/A+ Storyboard, in-image copy, design notes, image-generation Prompt, and evidence and compliance review. Suitable for Taobao, Tmall, JD.com, Pinduoduo, Douyin Shop, Xiaohongshu, Channels/WeChat Shop, as well as Amazon, Shopify, TikTok Shop, Temu, Shopee, and Lazada. Use when users need product images, detail pages, in-image copy, design briefs for designers, or localization for cross-border products. Not for short-video/livestream scripts only, ad placement or store operation optimization, or final legal conclusions requiring a lawyer (risk screening is possible); it does not automatically list or publish products. Do not fabricate qualifications, data, sales, reviews, certifications, or efficacy claims; verify official sources for current platform rules or high-risk compliance conclusions.
Package beat-level edit enhancement instructions from a B-roll plan, subtitle chunks, and optional style context. Use this when the goal is to turn B-roll matching into editor-ready guidance for keyword emphasis, micro-animation hints, A-roll stay-on-face logic, B-roll coverage style, and subtitle interaction
Generate professional presentation slides and high-quality illustrations using Gemini image generation API (Nano Banana 2), with interactive browser-based review and iterative editing. Full workflow: content planning conversation → slides_plan.json → batch image generation → review with feedback → targeted slide editing → PPTX packaging. Use when: user wants to create a presentation, make slides, generate a PPT/PPTX, prepare a talk deck, design visual slide content, or generate high-quality figures/illustrations for papers and documents. Do NOT use for: writing academic papers (use paper-writing) or planning academic conference talk narrative structure (use academic-slides).
This skill should be used when the user asks about "theme", "controlling idea", "thematic argument", "moral argument", "thematic arc", "flat arc", "negative arc", "the lie", "antagonist design", "motif", "symbolism", "theme audit", or wants to plan, track, or revise the thematic layer of a story, including the lie, truth, ghost-wound, and arc-type fields of a character file. NOT for finding or testing a story premise (use premise-workshop), or a character's profile, relationships, or "character arc" record (use character-management).
Animate objects in Decentraland scenes. Play GLTF model animations with Animator, create procedural motion with Tween, and chain sequences with TweenSequence. Use when the user wants to animate, move, rotate, scroll a texture, or create motion effects. Do not use for audio/video playback (see audio-video), player emotes (see player-avatar), authoring emote clips (see avatar-animations), or physics-driven motion (see player-physics).
Read real-time amplitude and 8-band frequency data from any AudioSource, AudioStream, or VideoPlayer entity in a Decentraland SDK7 scene with the AudioAnalysis component. Use when the user asks for music visualizers, beat reactivity, equalizers, or audio-reactive scenes. Do NOT use to play sound (see audio-video) or to detect player-emitted audio (this reads only entity-attached AudioSource/AudioStream/VideoPlayer audio).
Build a semantic inventory of local B-roll assets for edit planning. Use this when the goal is not yet to decide the final cut, but to understand what B-roll exists, what each asset proves, which ranges are usable, and how the asset library should be packaged for downstream matching.
Plan and design Decentraland games and interactive experiences. Use when the user wants game design advice, scene architecture, performance planning, or help structuring a game. Do NOT use for specific implementation (see add-interactivity, build-ui, multiplayer-sync).
Rewrite AI-generated text to sound natural and human-written. Removes LLM tells—cliché phrases, predictable structure, inflated language, and robotic patterns. Use when editing drafts, emails, articles, or any text that reads like it was written by AI
macOS screen capture, window recording, GIF conversion, and agent evidence bundles from the terminal. Built on ScreenCaptureKit for window-level targeting ffmpeg cannot do. Use when the user wants a screenshot of a specific window or app, a screen recording, a GIF conversion, a before/after diff, an evidence bundle for a PR, OCR text from a window, a terminal VHS recording, a Remotion render, or wants to watch a UI for changes. Requires macOS Screen Recording permission on first run.
Spotify music playback controls + a 5947-style database. Search, play, pause, manage queues, get scene-based music recommendations, and look up and explore styles. Use this skill when the user mentions: - Playback: “play music,” “play a song,” “put on some music,” “play,” “I want to hear XX,” “play some XX music” - Controls: “pause,” “resume,” “next track,” “previous track,” “volume,” “what’s playing” - Genre lookup: /genre, “look up music styles,” “recommend a music genre,” “what subgenres does XX have” - Scene/mood: “music for late night,” “energetic,” “ethereal,” “recommend music styles with XX characteristics” - Suno integration: using Suno to generate music without specifying a style - /spotify
Create complete stylized 3D animated shorts from a story idea through an ordered production workflow covering project brief, story outline, character and environment cards, standardized shot planning, text or optional pencil storyboards, video-model selection, single-shot generation, assembly, BGM matching, and final review. Use when the user wants an end-to-end narrative animation workflow with strong character consistency, scene continuity, timing, camera, performance, and audio control. Not for single images, simple edits, photorealistic live action, or one standalone clip.
Advanced rendering in Decentraland scenes. Billboard, TextShape, PBR materials, GltfNodeModifiers, and VisibilityComponent. Use when the user wants billboards, floating labels, 3D text, material effects, glow, transparency, or model node control. Do NOT use for screen-space UI (see build-ui) or loading 3D models (see add-3d-models).
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.