Media plugins

Add voice message transcription to NanoClaw using OpenAI's Whisper API. Automatically transcribes voice notes so the agent can read and respond to them. Triggers on "add transcription", "voice transcription", "whisper", "transcribe voice".

1 skill

Generate images locally with FLUX.2 Klein via mflux, on an Apple Silicon Mac. Free, offline, no API calls.

1 skill

Transcribe voice notes / audio files to text locally and offline (whisper.cpp, no API key) — Telegram .oga, wav, mp3, flac. FR/EN/auto.

1 skill

Turn an article into a 1-5 min narrated landscape video essay: Remotion + local TTS voice cloning + Pexels b-roll + kinetic typography. Fully local voice, no cloud TTS.

1 skill

Media processing pipeline — video transcription with WhisperX, audio extraction, and dependency management

1 skill

The master Remotion video skill, every format in one: a one-screen chooser (what the video is for, where it runs, the length, the look shown as moving samples), the story and script, then one engine for promos, demos, tutorials, explainers and more, with the brand motif in every transition, platform safe zones and measured verification.

1 skill

Promo videos with Remotion: bumpers, 15 and 30 s ads, teasers, launch films and landing page loops. Beat-locked, the brand's motif carrying the turn and the close, four aspects and RTL locales from one manifest. Single-purpose specialist of remotion-video-master, which covers every format.

1 skill

Product demo and product tour videos with Remotion: the real product, feature by feature, driven by a cursor with smooth zooms, voiced or caption-led, with cuts only inside screen recordings. Single-purpose specialist of remotion-video-master, which covers every format.

1 skill

Restore muffled or low-quality recordings: stem separation, EQ correction, high-frequency recovery.

1 skill

Generate images through the Codex CLI built-in image tool, billed via existing Codex auth instead of a separate API key.

1 skill

Pull the transcript of a YouTube video, with or without timestamps.

1 skill

Turns Claude Code into a motion-graphics studio: the film is written as code, rendered frame by frame in headless Chromium and encoded with ffmpeg, with a synthesized score (or optional ElevenLabs music) and a frame-critique loop.

1 skill

Creative director for agent-built video: settle the look, write a motion script, build and render with HyperFrames.

1 skill

Render 60fps 1080p MP4 animated videos deterministically with HTML5 Canvas/SVG, Playwright frame-stepping, contact-sheet visual self-review, Python audio synthesis, and FFmpeg.

1 skill

Edit raw action-cam fishing footage into YouTube edits, TikTok/Reels/Shorts catch features, full-trip recaps and 4K wildlife zooms.

1 skill

Archive YouTube videos with yt-dlp: download 480p video, transcript, description, metadata, thumbnail, and chapters. Chains fabric skills for analysis.

1 skill

Make short promo clips, trailers and comedy sketches out of a game's own engine: scripted acting, frame-by-frame renders with the game's sound, voices, synthesised effects, crash-in captions and an end card.

1 skill

Image generation via Gemini Imagen and friends.

Draft, schedule, and analyze posts across social platforms.

Kannaka Cannon — AI video intelligence via the clipcannon stdio MCP server (54 tools). A 22-stage pipeline ingests a clip, decomposes it into stems (scene/motion/speech/music/faces/OCR) that become HRM memories, and exposes editing, rendering, music/SFX generation, voice cloning, and lip-sync avatar tools.

A music player beside the chat, after the Meowzik Discord bot: YouTube, SoundCloud, Spotify and Suno, a queue, history and favorites with no limits, loop, shuffle, autoplay, volume and seeking. Claude can play and control it too, and it installs mpv and yt-dlp when they are missing.