图像与视频 Skill

---

name: create-music-elevenlabs

description: Generate an instrumental music bed via ElevenLabs Music, ROUTED THROUGH THE elevenlabs-proxy so it bills the Ads agent. Trims any sparse intro, loudnorm, fades the tail. Prompt + length from the template recipe. Use for the music layer of any video-ad format.

---

name: create-video-fal

description: Image-to-video (or text-to-video) via any FAL video model (Kling, Seedance, Veo), ROUTED THROUGH THE GooseWorks fal-proxy so the call bills the Ads agent. The template recipe names the model + params; image_url inputs must be public URLs (the orchestrator hosts local frames with the MCP media_upload). Returns the result video URL and downloads it. Use for the generative base clip of any video-ad format.

---

name: render-3d-character-explainer

description: Assemble an animated character-explainer video ad (any art style — Pixar-style 3D, claymation, anime, paper-cut, felt) (~77s, 9:16) built on an "N types of X" listicle spine — a recurring human protagonist plus a locked cast of N persona characters, one per list item. Given the per-scene i2v clips + a per-scene target-duration table + a narration track, it trims each clip to its scene window, re-encodes every segment to identical 1080x1920/30fps/libx264/yuv420p (decrease+pad, never crop) so the concat demuxer never drops frames, concats, and muxes audio — in RESTYLE mode the source ad's VO+music mix is reused verbatim, in ORIGINAL mode fresh per-scene VO (loudnorm I=-14) is mixed under an optional music bed (loudnorm I=-26). A static-still fallback loops a scene's keyframe when its clip is missing/failed, so the master always assembles; libass captions are burned last. FREE deterministic assembly (Python + ffmpeg, no bash, no paid calls); the recipe supplies the clips, keyframes, VO or source audio, and caption table and gates the paid cast-anchor/keyframe/Kling-i2v/VO/music calls to their own capabilities. Use for the 3d-character-explainer listicle format.

---

name: render-3d-product-showcase

description: Assemble a premium 3D product-showcase ad from a config — four beat clips (an orbiting hero rotation, a macro push-in, a physics reveal, a typographic close) normalized to the chosen backdrop colour, hard-concatenated in order, closed on a deterministic Playwright brand end card, and mixed under one instrumental bed at loudnorm I=-16 (music-only, no VO). Ships the runnable build_endcard.py + build_masters.py; the rotation/macro clips are create-video-fal i2v seeded on a create-image-fal styled hero, the reveal is Veo3 i2v, and the bed is create-music-elevenlabs. Use for the 3d-product-showcase format.

---

name: render-cartoon-music-video

description: Assemble a cartoon / animated / hand-crafted music-video ad from a config — a sung song carries the whole narrative while N per-bar i2v clips (one recurring animated character, one look pack) are each cut to their BAR window from librosa beat-tracking and hard-concatenated on the bar, VEED-whisper white bold-sans captions in the BOTTOM third (Alignment 2, above the logo bug, no pill) burned from the song's word timings re-spelled against the locked lyrics, a persistent brand logo bug held over the body (suppressed on the end card), and closed on a solid-brand-color PIL end card with the song still playing under it — never AI-rendered text. This is the FREE deterministic assembly stage (cut-to-bar + hard concat + logo bug + captions + end card + song mux); the song, character, keyframes, and clips come from create-music-elevenlabs / create-image-fal / create-video-fal. Use for the cartoon-music-video format.

---

name: render-cgi-sizzle

description: Assemble a 3D-CGI app sizzle — nano-banana CGI plates (blank-glow floating phone + a CGI studio with rim-light + placeholder burst shapes; studio look and phone finish come from the recipe choices) plus PIL compositing of the REAL App Store screenshots onto the bezel + burst-out overlays, driven by Kling 3.0 i2v steady-float per beat with a per-beat Ken-Burns FFmpeg push-in fallback when Kling garbles the UI, then VO/music mix (sidechain duck, loudnorm) + 1.15x speed + anti-AI grain finalize. The on-screen UI, any faces in it, and the wordmark are ALWAYS real assets composited via PIL — never AI-rendered. The paid steps (plates, i2v clips, VO, music) are separate capabilities; this ships the config + PIPELINE + FREE assembly and the recipe orchestrates the spend. Use for the cgi-app-sizzle video format.

---

name: render-cinematic-music-video

description: Assemble a cinematic live-action-style music-video ad from a config — an original sung anthem carries the whole narrative while N 35mm-film-look i2v clips are each cut to their lyric window and hard-concatenated on the beat as a 3-act arc, the anthem muxed at loudnorm I=-14, cinematic lower-third serif captions built from the song's OWN word timings (never Whisper) with the hook line landing on the chorus drop, and closed on a brand end card composited from the real asset — never AI-rendered text. This is the FREE deterministic assembly stage (cut-to-window + hard concat + anthem mux + captions + end card); the anthem, keyframes, and clips come from create-music-elevenlabs / create-image-gpt-image-fal / create-video-fal. Use for the cinematic-music-video format.

---

name: render-cosmic-mythology-voiceover

description: Assemble a cosmic-mythology-voiceover reel from a config — one spoken voiceover carries the whole narrative while N curated stills in one chosen look are weighted beat-synced across the delivered VO duration (cut_dur = VO_dur times weight over the weight sum, so emotional beats hold longer), Ken-Burns-zoomed per still (scale 2x, center crop, zoompan, fade-in first and fade-out last), ffmpeg-concatenated, the VO composited under the picture (libx264 crf18 plus aac), the ONE on-screen hook line faded on over the open with a drawtext alpha window, and Whisper/VEED captions burned along the bottom — never in-world text on a still. This is the FREE deterministic assembly stage (weighted sequence plus Ken-Burns plus concat plus VO composite plus hook overlay plus caption burn); the VO and the stills come from create-vo-elevenlabs and create-image-fal. Use for the cosmic-mythology-voiceover format.

---

name: render-flat-vector-explainer

description: Assemble the FREE steps of the flat-vector-explainer video format — a flat-illustration host character (who, where, tone, voice and music come from the recipe choices) walks a countable N-step product routine, one step per beat, and Remotion composites every chip/numeral/tagline/slate/CTA as an animated DOM overlay ON TOP of the Kling i2v character clips (text is NEVER baked into a keyframe — i2v warps type), the closing 'N products' grid is a PIL composite of the REAL product photos (not AI), full-sentence VO drives word-by-word burned captions over a VO-forward music bed, and the ~50s animated silent master is re-cut to a 30s deliverable FROM the animated master (never a static intermediate). Documentation-grade — ships config.example.json + PIPELINE.md + a README of the free assembly; the paid gen steps (keyframes, Kling i2v, VO, music) are separate capabilities the recipe orchestrates. Use for the flat-vector-explainer format.

---

name: bilibili-transcribe

description: 哔哩哔哩视频 → 下载 → 转录 → 存为 Markdown 的完整工作流。 支持 BV 号和完整链接，无需登录。

---

name: learning-notes-automation

description: 学习笔记自动化：视频/播客转录 → 知识点提取 → 闪卡生成 → 知识图谱更新。触发词：学习笔记、闪卡、Anki、知识提取、视频学习

---

name: weibo-transcribe

description: 微博视频 → 下载 → 转录 → 存为 Markdown。 支持 weibo.com 和 m.weibo.cn 链接，无需登录。

---

name: x-ingest

description: X(Twitter) 推文采集 → 统一 frontmatter Markdown。自动区分图文与视频： 图文下载图片本地嵌入，视频提取直链转成文字稿。无需登录（走官方嵌入端点）。

---

name: youtube-transcribe

description: YouTube 视频 → 下载 → 转录 → 翻译 → 存为 Markdown。 支持中英文，英文内容自动翻译成中文，输出中英对照。

---

name: zhihu-transcribe

description: 知乎视频 → 下载 → 转录 → 存为 Markdown。 支持知乎视频链接、问答内嵌视频和专栏文章，无需登录。

---

name: simulink-streaming-dsp

description: Building or editing Simulink (.slx) models that stream signals in frames using DSP System Toolbox — audio, vibration, radar, comms. Use when the prompt names a Simulink block (Discrete FIR Filter, Buffer, Unbuffer, Rate Transition, Downsample, Upsample, Sample-Rate Converter, FIR Decimation, Time Scope, Spectrum Analyzer, From Multimedia File) or an action (frame-based processing, InputProcessing, buffering, overlap, windowing, STFT, short-time Fourier, decimate, downsample, upsample, resample, multirate, anti-aliasing, tunable filter). Prevents silent numerical errors from sample-vs-frame mismatch, wrong block choice, or missing anti-aliasing.

---

name: skillpack-creator

description: Create a reusable SkillPack from a successful completed task. Use when the user wants to convert a one-off research, coding, analysis, or content workflow into a distributable local SkillPack with `skillpack.json`, local skills under `skills/`, starter prompts, start scripts, and an optional zip package.

---

name: audioaccessorykit

description: Support automatic audio switching for paired third-party Bluetooth headphones or earbuds with AudioAccessoryKit. Use when a companion app registers an audio accessory, an app extension reports worn/removed placement or connected source-device changes, or AccessoryControlDevice capabilities and errors need handling. Do not use for general AVAudioSession routing, Bluetooth transport, or initial accessory pairing.

---

name: avkit

description: Create media playback experiences using AVKit. Use when adding video players with AVPlayerViewController, enabling Picture-in-Picture, routing media with AirPlay, using SwiftUI VideoPlayer views, configuring transport controls, displaying subtitles and closed captions, or integrating AVFoundation playback with system UI.

---

name: callkit

description: Implement VoIP calling with CallKit and PushKit. Use when building incoming/outgoing call flows, registering for VoIP push notifications, configuring CXProvider and CXCallController, handling call actions, coordinating audio sessions, or creating Call Directory extensions for caller ID and call blocking.

---

name: dockkit

description: Control motorized camera docks and enable intelligent subject tracking using DockKit. Use when discovering DockKit-compatible accessories, implementing camera subject tracking for faces or bodies, controlling dock motors for pan and tilt, configuring framing behavior, setting regions of interest, or building video apps with automatic camera tracking.

---

name: manimce-best-practices

description: Trigger when: (1) User mentions "manim" or "Manim Community" or "ManimCE", (2) Code contains `from manim import *`, (3) User runs `manim` CLI commands, (4) Working with Scene, MathTex, Create(), or ManimCE-specific classes. Best practices for Manim Community Edition - the community-maintained Python animation engine. Covers Scene structure, animations, LaTeX/MathTex, 3D with ThreeDScene, camera control, styling, and CLI usage. NOT for ManimGL/3b1b version (which uses `manimlib` imports and `manimgl` CLI).

---

name: manimgl-best-practices

description: Trigger when: (1) User mentions "manimgl" or "ManimGL" or "3b1b manim", (2) Code contains `from manimlib import *`, (3) User runs `manimgl` CLI commands, (4) Working with InteractiveScene, self.frame, self.embed(), ShowCreation(), or ManimGL-specific patterns. Best practices for ManimGL (Grant Sanderson's 3Blue1Brown version) - OpenGL-based animation engine with interactive development. Covers InteractiveScene, Tex with t2c, camera frame control, interactive mode (-se flag), 3D rendering, and checkpoint_paste() workflow. NOT for Manim Community Edition (which uses `manim` imports and `manim` CLI).

---

name: matlab-use-scenario-builder

description: Generate driving scenes, scenarios, road surfaces, and 3D content from scenariobuilder.* sensor data (GPS, camera, lidar, actor tracks) using Scenario Builder for Automated Driving Toolbox. BUILD, EXPORT, or AUGMENT a virtual scenario/scene/map: ego or actor trajectories, trajectory smoothing, OpenCRG road-surface extraction, 3D asset generation, static-object placement, point-cloud georeferencing + elevation, lane-based ego localization, sensor-fusion tracking, scenario-event extraction (cut-ins, hard brakes, near-misses, ADAS disengagements), or export to RoadRunner, drivingScenario, OpenDRIVE, OpenCRG, OpenSCENARIO, or Unreal Engine. Also: log-to-scenario, scenario harvesting, accident/near-miss reconstruction, SOTIF (ISO 21448) and ISO 26262 scenario coverage, USGS-aerial-lidar augmentation, traffic-sign placement, vision-based vehicle classification for actor assets. NOT for raw-data import or multi-sensor sync/crop/offset/timestamp normalization — route those to matlab-import-driving-data.