Skills
96,770 packaged capabilities and playbooks for agents. Install any of them with one command.
Creative & media
Text-to-speech voice toolkit. DEFAULT ACTION: When called with text (e.g. /voicebox hello), IMMEDIATELY run: uv run $SKILL_DIR/scripts/voicebox.py generate \"Calm Narrator\" \"<text>\" --play. Do NOT ask questions, do NOT greet the user — just generate and play the speech. Also supports: voice cloning, multi-speaker conversations, recording, and transcription.
WorkRally CLI (workrally) — a full-workflow AIGC comic-drama video creation toolkit for AI agents. Supports AI image generation, AI video generation, video prompt optimization, canvas audio/music generation, Hunyuan 3D model generation, AI audio generation, full CRUD for projects/series/scenes/shots, asset library, media asset management, infinite canvas, file upload/download, and more. Use when user asks to generate images, generate videos, generate audio, generate music, generate 3d, optimize video prompts, manage projects, series, shots, upload files, download assets, manage materials, or interact with WorkRally platform via command line.
Hard-gate writing auditor. Use to independently audit the hard gates of a writing workflow before formally outputting the final article. Trigger scenarios: - a draft already exists and you are about to produce the formal final version - the project is located at `01_项目/内容创作/` - you need to verify artifacts such as Claim Ledger, Module Budget, Lint Report, Final Candidate Core capabilities: - independently check whether artifacts are complete - check whether form hard thresholds are exceeded - check whether claims match the body text - output `PASS / FAIL` and write it to `03-Reviews/Gate-Check-Report.md`
Extract, transcribe, and translate YouTube video transcripts using the YouTubeTranscript.dev V2 API. Supports captions, ASR audio transcription, batch processing (up to 100 videos), translation to 100+ languages, and multiple output formats. Use when working with YouTube videos, subtitles, captions, or video-to-text conversion.
Multi-platform account homepage video extractor — enter a platform name and an account id/link to automatically pull works from the homepages of four major platforms (Douyin/Kuaishou/Bilibili/YouTube), resolve download links, and support bulk downloads. Use cases: breaking down and frame-by-frame studying competitor account content, saving all works of a favorite creator to avoid deletions, bulk transferring material for re-editing, saving your own account's works as a local backup against loss, stocking up course/tutorial videos offline for repeated viewing, and quickly pulling a partner's works for due-diligence analysis before collaboration. Trigger words: 主页视频下载, 批量下载视频, 账号视频提取, 快手主页下载, B站视频提取, YouTube频道下载, 视频保存, 无水印下载, 竞品视频下载, 博主视频备份.
Run 250+ 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
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.
This skill should be used when the user asks about "production management", "film budget", "scheduling", "crew management", "location logistics", "risk assessment", "act as a production manager", "production manager mode", "line producing", "production budget", "crew roles", "film insurance", "production timeline", "location permit", "production paperwork", or needs expertise in film budgets, production scheduling, crew management, location logistics, and risk mitigation for film projects. Part of the AlterLab FC Skills collection (Cinema & Digital Media department).
This skill should be used when the user asks about "audio production", "ElevenLabs", "voice isolator", "audio post-production", "AI narration", "text to speech production", "voiceover studio", "audio native", "transcription", "Scribe", "multi-track audio", "audio assembly", "batch audio processing", "audio export", "act as an audio producer", "audio producer mode", "TTS production", "podcast audio", "audiobook production", "narration workflow", "content series audio", "multi-tool audio chain", "ElevenLabs Projects", or needs expertise in end-to-end audio production pipelines using ElevenLabs tools. Part of the AlterLab FC Skills collection (GenAI pack).
This skill should be used when the user asks about "image-to-video", "still to motion", "animate my image", "Higgsfield video", "Soul Cinema Preview", "Seedance 1.5 Pro", "AI video generation", "Sora 2", "Kling 2.6", "Kling 3.0", "Veo 3.1", "Wan 2.6", "MiniMax Hailuo 02", "Higgsfield Assist", "Soul Cast", "content scoring", "act as an image-to-video director", "image-to-video mode", "camera motion on stills", "dolly on image", "Soul ID video", "character consistency in video", "upscale video", "video from photo", "motion intensity", "video duration control", "multi-shot storytelling", "social media video formats", "9:16 video", "16:9 video", "4K upscale", or needs expertise in converting still images into motion video using Higgsfield's image-to-video pipeline, camera motion application, Soul ID character tracking, and multi-shot narrative sequencing. Part of the AlterLab FC Skills collection (GenAI pack).
This skill should be used when the user asks about "multimedia storytelling", "interactive story", "longform web feature", "cross-platform narrative", "scrollytelling", "immersive journalism", "act as a multimedia story builder", "multimedia story mode", "web documentary", "transmedia narrative", "visual storytelling", "digital longform", "parallax story", "Shorthand", "Scrollama", "StoryMap", "interactive feature", "Juxtapose", or needs expertise in designing and producing cross-platform narrative experiences for the web. Part of the AlterLab FC Skills collection (New Media & Communication department).
This skill should be used when the user asks about "podcast production", "episode planning", "podcast script", "show notes", "audio editing", "podcast launch", "RSS feed", "act as a podcast producer", "podcast producer mode", "interview preparation", "podcast format", "podcast distribution", "recording setup", "podcast monetization", "podcast microphone", "DAW workflow", "audio mastering", "podcast hosting platform", "episode outline", "podcast trailer", "podcast editorial calendar", or needs expertise in planning, scripting, producing, and distributing professional podcast episodes. Part of the AlterLab FC Skills collection (New Media & Communication department).
Higgsfield image prompt director for Banana Pro, Soul Cinema, and GPT-2. Modes: (0) face lock for new characters — Banana Pro single-pass (default), GPT-2 single-pass (higher fidelity, higher credits), or Soul Cinema two-pass, all on mid-gray seamless (locked default; white on request) with black camisole/tank baseline, (1) single-image character outfit — Banana Pro or Soul Cinema two-step, (2) 6-panel character sheet, (3) cinematic scene plates with or without characters, (4) GPT-2 detail face/chest-up, (5) outfit replacement — swap a face onto an outfit using two refs. Reads references for hair, makeup, wardrobe, jewelry, identity. Outputs photorealistic prompts with one clean cinema stack — pores, subsurface scattering, strand hair, fabric weave, atmospheric perspective, anamorphic character, theatrical grain. Use for new character builds, character/outfit refs, character sheets, scene plates, environment plates, detail shots, outfit replacement, or any photorealistic still.
The 边淬火 2.0 long-form creation workflow. Used to write 公众号/Substack/long-post articles on 2.0 themes such as AI, business, making money, one-person businesses, content monetization, market feedback, product hypotheses, and business games. You must first read the local content-management and content-benchmark library nodes, expand the surrounding nodes, and then assemble the logic chain; parallelism, parallel synonymous sentence patterns, empty aphorisms, and writing without reading the libraries are prohibited.
Bilibili search bulk download tool. Enter a keyword to search videos across Bilibili, with sorting by play count, likes or time; after searching, automatically fetch all HD download links in bulk. Use when users need to find Bilibili videos, search Bilibili content, download Bilibili videos or collect Bilibili material. Trigger words: B站搜索, bilibili搜索, B站下载, B站视频下载, b站搜视频, B站找素材, 下载B站, bilibili下载.
Blender 5.x modifiers, bmesh API, mesh editing operators, sculpting setup — SubSurf, Boolean, Array, Mirror, Bevel, bmesh procedural mesh creation, and modeling pipelines via Python (bpy). Includes 5.1 changes (Boolean speed, Corrective Flip Normals, UV improvements).
Use the CreatiBI CLI (cbi) to manage material libraries, projects, and project sets, including files, folders, tags, associated products, video understanding signals, AI video analysis results, highlight clips, project lists and creation, project script/material lists, project set lists and project lists under project sets; use when the user mentions uploading to a material library, querying files or projects, viewing signals or storyboards, maintaining material metadata, viewing project sets, or initializing cbi or logging in before operating.
Tools, patterns, and utilities for creating music with code. Output as a .mp3 file with realistic instrument sounds. Write custom compositions to bring creativity to life through music. This skill should be used whenever the user asks for music to be created. Never use this skill for replicating songs, beats, riffs, or other sensitive works. The skill is not suitable for vocal/lyrical music, audio mixing/mastering (reverb, EQ, compression), real-time MIDI playback, or professional studio recording quality.
Color grade clips the way a professional colorist does — normalize, balance, match, then build a look — using inspect_color scopes and apply_color (exposure, white balance, wheels, curves, hue curves, LUTs). Use when the user asks to color, grade, match shots, fix white balance, fix exposure, neutralize a cast, or apply a look/film stock.
Co-Write mode for Chinese web fiction. Use when the user wants to write together with AI instead of handing over a whole chapter; is stuck on what to write next; wants AI to continue from their own writing; wants several versions in different directions (recommended + alternatives) to choose from; or wants to set only the goal, conflict, and emotion and have AI write the prose. Three modes: Auto (AI writes automatically), Assist (I write some, AI fills in some), and Director (I set the direction, AI produces multiple versions).
Break down what a video, image-text post, or piece of copy actually provides in terms of value, from the audience's perspective. Use the five-dimension model of cognition, emotion, identity, action, and social to analyze the whole, key nodes, or sentence by sentence; use when the user says "break this one down", "audience-perspective analysis", "what value does this give the audience", or "break it down sentence by sentence".
The third step of the content creation pipeline: polishing a judgment seed into shape — the trio (common-sense stance → reversal → mechanism) plus the nail. The AI performs differential diagnosis, writes prescriptions, presses with follow-up questions to deliver it, and challenges from the opposing side. Not a single word of the judgment's body text is ghostwritten. Trigger: /cuihuo-judgment, "help me polish this judgment", "shape the judgment", "polish the judgment"
Comprehensive design skill: brand identity, design tokens, UI styling, logo generation (55 styles, Gemini AI), corporate identity program (50 deliverables, CIP mockups), HTML presentations (Chart.js), banner design (22 styles, social/ads/web/print), icon design (15 styles, SVG, Gemini 3.1 Pro), social photos (HTML→screenshot, multi-platform). Actions: design logo, create CIP, generate mockups, build slides, design banner, generate icon, create social photos, social media images, brand identity, design system. Platforms: Facebook, Twitter, LinkedIn, YouTube, Instagram, Pinterest, TikTok, Threads, Google Ads.
Turn a novel into a Word draft for editing. Use when the user wants to export a manuscript to Word/docx, organize it into chapters with a table of contents, left-align chapter headings, mark sections for them to revise, or says, "Write a rough version first; I'll select and revise it." Produces a Word draft (table of contents + left-aligned Chapter X headings + red-highlighted 【修改】 suggestions), named 《书名》修改N, and never overwrites the original.
Writing English web fiction for international platforms (Wattpad / Royal Road / Webnovel / Radish / KDP). Use when the user wants to write an English novel, publish on English-language platforms, adapt a Chinese story into natural English (not translationese), ask how to write for or choose an English-language platform, or remove “Chinese web novel translationese / AI-like” phrasing from English. Produce English that reads like it was written by a local author, following English web fiction conventions.
World-class explainer video expertise combining the visual storytelling of Kurzgesagt, the simplification genius of Common Craft, and the conversion science of high-performing SaaS landing pages. Explainer videos transform complex ideas into clear, compelling stories. Great explainers don't just inform—they illuminate. They take abstract concepts and make them feel obvious. The best explainer videos understand that simplicity is the ultimate sophistication, that showing beats telling, and that the goal isn't to impress with complexity—it's to make the viewer feel smart. Use when "explainer, explainer video, how it works, product demo, onboarding video, training video, educational video, whiteboard, animated explainer, explain this, simplify, concept video, SaaS video, explainer, video, animation, educational, product, demo, onboarding, SaaS" mentioned.
Turn article passages, technical concepts, or short explanations into 16:9 white-background, mostly black-and-white, hand-drawn line illustrations using Codex image generation/image-2. Use when the user wants a passage, section, or paragraph in Chinese, English, or mixed language transformed into visual explanation shots, methodology diagrams, process/state/action illustrations, or a shot list where a default hand-drawn rabbit IP participates in the core cognitive action rather than decorating the scene.
AI script supervisor for continuity, dialogue tracking, script notes, scene analysis. Use when: tracking continuity, script supervision, dialogue notes, scene analysis, production reports. Keywords: film, script, continuity, dialogue, continuity.
Author or modify a 3D CAD part for Forgent3D with one CLI, cad-build. Two routes, picked from the request and the files present — a parametric model.py in the build123d dialect, built on the bundled wasm kernel, measured (cad-build rebuild / check), rendered (snapshot), exported to STEP/STL/GLB and published as a shareable 3D link (preview); or an OpenSCAD .scad, rendered by the bundled OpenSCAD (rebuild reads bbox, volume and echo()), snapshotted, its Customizer parameters walked (knobs), exported to STL/3MF, and handed off as a link that carries the code and converts it to STEP. Use whenever the task is creating or changing a 3D CAD part or assembly, previewing or converting a .scad, or a model.py in the build123d dialect or a .scad file is present.
Plan and design Decentraland games and interactive experiences. Scene limit formulas, performance budgets, texture requirements, asset preloading, state management patterns (module-level, component-based, state machines), object pooling, UX/UI guidelines, input design, and MVP planning. 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).
Selects and prototypes player-facing mechanics for meaningful choices, challenge, progression, quests, combat, collection, rewards, and tutorials. Use for deciding what the player repeatedly does and how that action supports the intended experience; defer numerical tuning to game-balance and cross-system feedback analysis to game-systems-design.
Short-form writing for WeChat official accounts (under 1,000 characters, text only, no images). Turn an observation, a test, or a line of thought into a quick, punchy post, or compress a long article or raw material into a short piece. The core is a set of style rules: first person without preaching, no AI tone, rhythm through paragraph breaks, and endings that land on a concrete image. Triggered when the user says "write something short", "official account short post", "under a thousand words", "jot one down", "write a short piece about this", "compress into a short post", or "the no-images kind".
Apply narrative structure and storytelling techniques for brand, business, and communication contexts. Use this skill when the user needs to craft a compelling story, build a brand narrative, structure a presentation around a story arc, or use storytelling to communicate data or strategy — even if they say 'tell a better story', 'make this presentation more engaging', or 'how do we build our brand story'.
All-in-one image studio. One Skill handles four high-frequency image types: Xiaohongshu covers (3:4), PPT illustrations (16:9), various charts (flowcharts/architecture diagrams/heatmaps and 12+ other types), and article logic diagrams (16:9). Built-in 12 visual styles + automatic style recommendation, covering Chinese-language scenarios. Flow: enter content → choose type → fill in details → choose style → prompt (automatic / review) → generate image. Calls LabNana's GPT-image-2 model and outputs PNG. Trigger words: "make a cover", "Xiaohongshu cover", "make an illustration", "PPT image", "draw a flowchart / architecture diagram / heatmap / mind map", "article logic diagram", "logic diagram", "add an image", "image-studio".
Use when the user asks to generate or edit images via the OpenAI Image API (for example: generate image, edit/inpaint/mask, background removal or replacement, transparent background, product shots, concept art, covers, or batch variants); run the bundled CLI (`scripts/image_gen.py`) and require `OPENAI_API_KEY` for live calls.
Elite frontend image-direction skill for generating premium, conversion-aware website design reference images via ChatGPT image generation. Fires when the user asks to generate website mockups, landing page concepts, section design references, or UI comp images. Enforces ONE separate horizontal image PER section, composition variety (bans the default left-text/right-image on every section), background-image freedom, varied CTAs, varied hero scales, narrative concept spine, second-read moments, and a single consistent palette across all images. Outputs structured prompt blueprints that produce Awwwards-tier visual references a developer or coding model can accurately recreate. Image generation only - does not write code.
Edits raw travel footage into polished Instagram Reels (9:16, 1080×1920) with cinematic color grading, beat-synced cuts, text overlays, transitions, and background music. Use when the user mentions editing a reel, travel reel, Instagram video, cutting footage, travel clips, reel from footage, or provides folders with destination subfolders and optional shared _music or _fonts.
Invokes Google Gemini models for structured outputs, image generation, text-to-speech narration, multi-modal tasks, and Google-specific features. Use when users request Gemini, image generation, Gemini TTS or a synthesized voice, structured JSON output, Google API integration, or cost-effective parallel processing.
Automatically turns a single subject concept (Chinese or English, such as "sound phenomena", "lever principle", "photosynthesis") into dynamic teaching content. Both outputs are rendered by HyperFrames and share the same index.html (switched by the mode variable): ① narrated teaching video — storyboard + Minimax TTS Chinese narration + subtitles + full MP4 (1080p, ~90s, for posting on video channels/showing to kids); ② silent looping animation — a compact silent MP4 of the same content (1080p, ~32s, looping, autoplaying when embedded in the body of a WeChat official account article). Both share one 7-segment storyboard and the same theme palette (sound/light/force/electricity/heat/biology/math each have their own color scheme), so the video and the animation have a completely unified style. Use this skill when the user mentions "teaching video", "narrated video", "teaching animation", "silent animation", "looping video", "animation embedded in a WeChat official account article", "concept → animation", or gives a subject concept to be turned into dynamic content.
Create, revise, audit or polish submission-grade manuscript figures for Nature-family and other high-impact journals using Python or R. Covers multi-panel scientific plots, matplotlib/seaborn, ggplot2/patchwork/ComplexHeatmap and journal-ready SVG/PDF/TIFF exports. Also use for OpenRouter GPT Image 2 graphical abstracts, concept or mechanism schematics when the user explicitly requests AI generation. Establish the figure conclusion, evidence logic and exports before execution; honor the chosen or saved plotting backend. Trigger on paper figures, scientific/academic plotting, manuscript data visualization, 论文配图、学术写作配图、科研绘图、科研作图、画图、作图、出图、论文图表、可视化、论文示意图、机制示意图、图形摘要. Not for dashboards or Illustrator/Figma-first infographics.
Top-level fiction writing and review workflow for using distilled novel styles without turning outlines, settings, rules, or procedures into lifeless prose. Use when writing, rewriting, outlining, or reviewing fiction with novel-style skills; enforces story-engine checks before use-novel-style and requires eval-novel-style for style review.
Convert and compress media from your agent with OpenClip — video to mp4, webm, mov, mkv, or gif, and audio to mp3, aac, wav, or flac. FREE with just an OpenClip account, no subscription. Use when asked to “convert this to mp4”, “turn this into a gif”, “extract the audio as mp3”, “make this a webm”, “rip the audio”, “this file is too big”, or “change the format”.
Pull still frames out of a video as thumbnail images with OpenClip, then compress, resize, crop, or reformat them to jpg, png, or webp. FREE with just an OpenClip account, no subscription. Use when asked for “a thumbnail from this video”, “a still frame”, “a poster image”, “preview frames”, “a contact sheet”, “grab a screenshot from the video”, or “resize this image”.
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.