Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Creative & media
Design and build the visual layer of a long-form talking-head video as one persistent world the camera moves through, instead of a sequence of separate graphics. Use when asked to edit a video, add motion graphics, design the visuals for a script, plan an edit, or fix an edit judged confusing, text-heavy, hard to follow, or "clearly made by an AI". Covers the persistent world, camera grammar, the spotlight rule, spatial open loops, ornament and token discipline, transcript timing, and the verification gates.
Local vision CLIs: glance (describe/ask/OCR an image), ground (locate a target, pixel box), detect (element inventory), trace (image to SVG geometry), crop (cut a pixel box to a file), and scripts/html_shot.py (HTML file to image). Use for any task involving an image — questions, text, splitting and transcribing long screenshots or chat histories, locating elements, comparing, rebuilding as HTML/SVG, digitizing a sketch or diagram, reading values off a chart, operating a GUI from screenshots — and to re-check an image yourself when a description you were given lacks a detail.
Babylon.js 8 3D engine development expertise with curated API patterns, code examples, and on-demand documentation access. Use when working with Babylon.js scenes, meshes, materials (PBR/Standard), cameras, lights, shadows, GUI (2D/3D), animations, physics (Havok), thin instances, glTF loading, post-processing, WebXR, procedural/code-built 3D models, or any 3D rendering task. Covers: (1) Scene setup and engine initialization (WebGL/WebGPU), (2) Mesh creation, transforms, instancing, and merging, (3) PBR and Standard materials with textures, (4) Camera types (ArcRotate, Universal, Follow), (5) Lighting and shadows, (6) 2D/3D GUI with AdvancedDynamicTexture, (7) Animation system and animation groups, (8) Asset loading (glTF, OBJ, STL), (9) Performance optimization and profiling, (10) ASRS/warehouse digital twin visualization patterns, (11) Procedural modeling: building high-quality 3D models from primitives with animation-ready parent/pivot hierarchies, CSG2 booleans, lathes, extrudes, and custom VertexData.
Establishes a brand's visual direction and compresses it into a Creative Direction Brief with a paste-ready 'style block' that every downstream image prompt reuses, so all generated assets look like they came from one brand. Use when the user says 'define my brand style', 'I want a consistent look', 'what should my visuals look like', 'help me pick an aesthetic / mood / color palette', 'my generated images all look different', or before any multi-asset creative work where no style guide exists yet. Produces the Creative Direction Brief: aesthetic profile, palette with hex codes, style block for prompts, and prohibitions.
(macOS, requires schedule-manager) Use this skill whenever the user wants to write a personal diary entry or daily journal — this includes any request to record today's events, write a diary, log what happened today, or capture personal reflections. Invoke immediately for phrases like 帮我写日记, 写日记, 记录今天, 今天的日记, "write my diary", or "daily log". This skill guides a complete journaling session: reviewing today's tasks from Reminders, reflective guided questions, composing the entry, and scheduling follow-up plans. Distinct from diary-note (quick append) and weekly-report (work summary).
Create and manage script- or audio-driven AI avatar videos from an official or saved presenter, an imported portrait, or a representative still prepared from video. Use for 数字人、数字分身、虚拟人、照片开口说话、人像口播、改稿不重拍, AI avatar, AI avatar video, avatar video, talking avatar, talking photo, photo avatar, video avatar, AI presenter, virtual presenter, virtual spokesperson, digital human, or a person's digital twin, including choosing or creating the avatar and deciding its voice, script, aspect ratio, and output quality. Do not use for a static profile-picture avatar, game or 3D character creation, an industrial digital twin, ordinary B-roll, generic image-to-video, video translation, talking-head editing, lip-sync dubbing alone, or TTS-only requests.
Complete fal.ai text-to-video system. PROACTIVELY activate for: (1) Kling 2.0/2.5/2.6 Pro video generation, (2) Sora 2 for creative videos, (3) LTX Video with audio, (4) Runway Gen-3 Turbo for fast iteration, (5) Luma Dream Machine, (6) Video duration and aspect ratio, (7) Motion prompt engineering, (8) Camera movement keywords. Provides: Model endpoints, quality tiers, prompt structure, duration options. Ensures cinematic video generation with proper motion description.
Complete audio visualization system. PROACTIVELY activate for: (1) Animated waveforms (showwaves), (2) Static waveform images (showwavespic), (3) Spectrum analyzers (showspectrum), (4) Frequency bar visualizations (showfreqs), (5) Stereo vectorscope (avectorscope), (6) Musical note display (showcqt), (7) SoundCloud-style waveforms, (8) Music video visualizers, (9) Podcast waveform videos, (10) Combined visualization dashboards. Provides: Filter parameter tables, color scheme options, scale comparisons, template commands for music videos and podcasts. Ensures: Professional audio visualizations for content creation.
Discover music, get personalized recommendations, and download high-fidelity audio files. Use when user wants to find new music based on their taste, search for songs/albums/artists, get recommendations similar to artists they like, or download lossless audio (FLAC/Hi-Res) from Qobuz or TIDAL. Trigger phrases include "find music like", "recommend songs", "download album", "lossless", "Hi-Res", "FLAC", "music discovery", "similar artists", "setup music".
Create a distinctive personal website through a phased reference-image-driven process: locking content and style, locking reference images, confirming asset breakdown, reviewing asset generation, then high-fidelity replication and QA. Suitable for building, redesigning, or replicating personal websites, portfolios, creator sites, founder sites, or personal IP sites using Codex, Image 2, reference images, generated assets, or a high-quality image-to-code workflow.
Generate images with Gemini (default) or fal.ai FLUX.2 klein 4B (--cheap for fast/low-cost). Generate videos with Grok Imagine (default) or fal.ai LTX-2 (--cheap). Use for: create image, generate visual, AI image generation, poster, video generation.
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.
Generate AI videos using the Pollo AI API. Supports 13 leading models (Kling, Sora, Runway, Veo, Pixverse, Hailuo, Vidu, Luma, Pika, Wan, Seedance, Hunyuan, Pollo) with 50+ versions. It also supports task polling, credit cost estimation, and credit balance checks. Use this skill whenever the user wants to generate an AI video from text or image, use any AI video model, check Pollo credits, or mentions Pollo AI, pollo.ai, or any of the supported model names. Even if the user just says "generate a video" or "make me a short clip" without mentioning Pollo, this skill should be used.
Unity 6 2D game development guide. Use when building 2D games, working with sprites, sprite atlas, SpriteRenderer, tilemaps, 2D physics (Rigidbody2D, Collider2D), 2D lighting, sorting layers, or sorting groups. Based on Unity 6.3 LTS documentation.
Generate a Chinese-narrated recap film or an original-audio narrative short from an input video. Use when the user provides .mp4 / .mov / .mkv / .webm and asks for editing, adding narration, dubbing, summarizing, or narration for short dramas/TV series/movies/documentaries/science explainers. Orchestrates the video-* skill chain: video understanding → Agent designs the story and audiovisual plan → editing → dubbing → compositing. Triggers: video explanation, video narration, generate recap, video recap, voiceover, narration, auto-dub, recap.
Synthesize Chinese voiceover audio from a timestamped narration.json. Generates speech segment by segment using MiMo TTS (mimo-v2.5-tts), Fish Audio (s2.1-pro-free), or an explicitly configured generic IndexTTS HTTP service, dynamically adapting speech rate to time windows and handling loudness; takes narration on the input timeline and produces tts_segments and tts_meta.json on the output timeline. Data disclosure: each narration segment's text is sent to the chosen TTS service (MiMo / Fish Audio / user self-hosted IndexTTS), and the --voice-ref reference audio is sent to MiMo. Also includes an experimental English-to-Chinese dub path: runs only when dub mode is explicitly selected and --confirm-voice-rights is passed; it sends the source video audio to MiMo ASR and clones the voice using the original speaker's voice as reference via MiMo voiceclone; use only for content you have the right to use and whose speaker has consented to voice cloning. Trigger words: 配音, 语音合成, TTS, 解说配音, voiceover, text to speech, 旁白配音.
Create Xiaohongshu covers in Yetong’s original live-action style for fitness coaches serving local, in-person members. Use when there is a specific topic or draft and a cover, prompt, numbered composition selection, further refinement, or direct generation is needed. After reading the user’s activated DNA, follow the original step-by-step, one-question-at-a-time process, with five fixed compositions, a 3:4 aspect ratio, oversized text at the top and bottom, and a white, warm orange-yellow visual style. Not for broad topic ideation, changing only the posting title, or reviewing an entire draft.
Yu Yile's primary AI video prompt Skill. First choose a tier based on the script style, then break a script, story, voiceover, storyboard, or even a single idea into directly generatable video prompts, like a director. The short-video tier (Douyin/Xiaohongshu pacing: one-take shots, multiple shots, street photography, character vignettes, vertical short dramas; benchmarks include 《丹眉》 and Cyber Atomic) and the cinematic tier (slow-paced stories, acting, dialogue, scenes between characters, AU micro-expressions, and the eight-dimensional formula) each have their own structure and examples. Each segment includes generation mode and parameters, reference-image binding, characters, props and setting, shot timeline, audio, and hard constraints. Supports Seedance 2.0/2.5, 即梦, dreamina, LibTV, and TT AI. Use when users want video prompts, AI video prompts, Seedance prompts, 即梦 prompts, script-to-video conversion, shot breakdowns, storyboard prompts, fast-cut edits, short-video prompts, a storyboard, a shot list, multiple image references, first and last frames, character performance prompts, AI acting, emotional progression, director-style prompts, or want a script/story/idea turned into a video. If the current project has a dedicated workflow Skill (Vlog, emotional hooks, news narration, etc.), use that first.
Image generation skill, to be used when the user needs to generate images, visual infographics, create images, or edit/modify/adjust existing images. Official image generation service based on the ChatGPT Image 2 model (gpt-image-2) on the API易 platform (https://api.apiyi.com/). The model supports precise size/quality control (including 4K) and is billed by token. Key differences from gpt-image-2-all (the unofficial reverse-engineered version): uses the /v1/images/generations and /v1/images/edits endpoints; has an explicit size parameter; has a quality parameter; is billed by token; uploads reference images with multipart/form-data; b64_json is plain base64 with no prefix.
Agent 2 of the video editing pipeline. Finds and removes spoken mistakes — stutters, repeated words, false starts, and retakes (re-recorded lines) — from a talking-head recording. Use after cut-silences, when asked to cut mistakes, remove stutters / repeats / filler restarts, clean up flubs, or keep the best take. Works review-gated: it proposes every cut with context and a reason for approval, then renders only the approved cuts via ffmpeg. Requires a word-level transcript.
Generate realistic audio from text using ElevenLabs Text-to-Speech API. Use when the user needs to convert text to speech, create voiceovers, generate narration, or produce audio content. Triggers include "generate audio", "text to speech", "TTS", "voiceover", "narration", "ElevenLabs", "audio from text", "read this text aloud".
Complete fal.ai image-to-image system. PROACTIVELY activate for: (1) FLUX image-to-image transformation, (2) ControlNet (canny, depth, pose), (3) Inpainting with masks, (4) Upscaling (ESRGAN, Clarity), (5) Background removal (BiRefNet, RemBG), (6) Face restoration (GFPGAN, CodeFormer), (7) IP-Adapter style transfer, (8) Strength parameter tuning. Provides: Transformation endpoints, mask formats, ControlNet types, upscaling options. Ensures proper image editing and enhancement workflows.
Use the GLM vision MCP for image understanding and analysis. Trigger conditions: (1) the user asks to analyze an image, understand an image, or describe image content (2) objects, text, or scenes in an image need to be recognized (3) the GLM visual understanding capability is to be used
Use this skill whenever the user wants to use Grok or xAI through the grok-cli command line, including chat, X search, image generation/editing, video generation/editing/extension, text-to-speech, speech-to-text, OAuth login/status, or local usage stats. This skill should also be used when grok-cli may need to be installed first; it runs the user's requested grok-cli command directly when possible, repairs install/auth/permission problems only after an actual failure, retries the original task after recovery, and returns Grok/grok-cli results without assistant-side rewriting unless the user explicitly asks for transformation.
WeChat Official Account investment blogger distiller. From Official Account article data, it automatically extracts a seven-dimension DNA profile — trading system, market judgment, expression style, content depth, engagement features, hot-topic map, and persona genes — and outputs a structured style profile. Trigger words: distill investment blogger, distill influencer, distill Official Account influencer, analyze investment blogger style, extract trading system, style profile, investment blogger distillation.
Run an iterative audit-and-fix loop on Thai prose using the kien-thai skill — repeats audit/fix passes until a full pass produces zero new edits. TRIGGER when user invokes `/kode-thai`, asks for "audit loop" / "วน audit" / "ตรวจวนๆ" / "ขัดภาษาไทยให้สุด" on Thai writing, says variants of "แก้ไปเรื่อยๆ จนกว่าจะไม่เจอที่ผิด", or explicitly requests repeated review passes on Thai prose. DO NOT TRIGGER for single-pass rewrites or one-off Thai edits (use kien-thai directly), or for non-Thai content.
Work with Lapian Notes / 拉片笔记 (github.com/bkingfilm/lapian-notes) — a local-first React/Vite tool that turns a film into an editable shot-by-shot study notebook: local frame extraction, AI-assisted structure analysis (bring your own AI, no API key required), story-line swimlane timeline, structure tree, and audience-emotion curve. Use when the user asks about Lapian Notes, "拉片笔记", "拉片" (shot-by-shot film analysis) tooling, cloning/running this repo (npm run dev, run.bat/run.command), the AI-analysis-package (ZIP) round-trip workflow, or contributing a PR to lapian-notes. Not for generic video editing (use `opencut` for that) or generic film-analysis theory unrelated to this codebase.
Video/audio/image processing with FFmpeg and ImageMagick. Tools: FFmpeg (video/audio), ImageMagick (images). Capabilities: format conversion, encoding (H.264/H.265/VP9/AV1), streaming (HLS/DASH), filters, effects, thumbnails, watermarks, batch processing, hardware acceleration (NVENC/QSV). Actions: convert, encode, resize, crop, compress, extract, merge, stream, transcode media. Keywords: FFmpeg, ImageMagick, video encoding, audio extraction, image resize, thumbnail, watermark, HLS, DASH, H.264, H.265, VP9, AV1, codec, bitrate, framerate, resolution, aspect ratio, filter, overlay, concat, trim, fade, batch processing. Use when: converting video/audio formats, encoding with specific codecs, generating thumbnails, creating streaming manifests, extracting audio from video, batch processing images, adding watermarks, optimizing file sizes.
Bloom and glow effects using Three.js UnrealBloomPass with React Three Fiber. Use when implementing glow, bloom, luminance-based effects, selective bloom on specific meshes, or neon/ethereal lighting. Essential for cyberpunk aesthetics, energy effects, magic spells, and UI glow.
Use when someone wants seamless looping motion graphics in their own brand colors: animated backgrounds behind titles, filler for the dead seconds between AI shots, bumpers, transitions, or shape animations, built as code and rendered to MP4, optionally with their mascot, product, or character from Scenario composited over the loop. Keywords: seamless loop, motion background, canvas animation, motion graphics, loop seam, motion blur, H.264 export, alpha composite
WeChat official account topic and outline planning skill. Turns a rough idea, resource pack, voice draft, or interview transcript into 2-3 high-value topic angles, 1 recommended direction, 1 main outline, and 1 backup outline. Suited to the pre-writing planning stage: confirm the structure first, then hand off to the drafting skill.
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.