chatcut
ArtCog/chatcut
One command turns raw talking-head footage into a finished video — driven by your coding agent (Claude Code / Codex). Orchestrates video-use, hyperframes, auto-editor, faster-whisper, ffmpeg + hard-won correctness rules.
Instalação
npx convez-ai ArtCog/chatcut1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
README
A tradução para o português deste README ainda está na fila.
chatcut
One command turns raw talking-head footage into a finished video — driven by your coding agent (Claude Code / Codex), not a timeline.
chatcut is an extensible, agent-orchestrated video-editing studio focused on talking-head content (vlogs, tutorials, explainers). You drop in raw footage, say what you want in plain language, and the agent runs a battle-tested pipeline: transcribe → cut pauses & stumbles → subtitles → motion graphics → color → render.
It does not reinvent the engines. It orchestrates the best open tools — video-use, hyperframes, auto-editor, faster-whisper, ffmpeg — and adds the part that's actually hard: opinionated, hard-won correctness rules so a single command doesn't produce a broken video (no frozen frames, no A/V desync, no crushed audio).
🚧 Status: v0.1, building in the open. The core pipeline works end-to-end (raw → finished video in one command); polishing setup and real-world testing.
Quick start
git clone https://github.com/ArtCog/chatcut && cd chatcut
./setup.sh # Windows: ./setup.ps1 (installs chatcut + free local toolchain)
chatcut tools # see capabilities and what's ready
chatcut edit raw.mp4 # raw footage → projects/raw/renders/final.mp4
Then just talk to your agent: "edit raw.mp4 — cut the filler, add captions, warm look." It runs the pipeline, shows you a cut-plan, renders a preview, and finalises.
Why chatcut
- One command, not copy-paste. A real CLI + agent orchestration, instead of pasting prompts into an IDE.
- Talking-head depth. Cut by meaning, kill filler, fix stumbles, sync captions to words.
- Free by default. Local
faster-whisperout of the box — no paid API key required. Premium engines (ElevenLabs Scribe) are an opt-in upgrade. - Cross-platform encode. NVENC when available, graceful fallback to libx264 / VideoToolbox / QSV.
- Built to grow. A plugin/tool registry — new tools and pipelines drop in without touching the core.
How it works (talking-head pipeline)
raw footage
→ normalize (clean CFR) don't desync on variable-framerate input
→ transcribe (word-level) faster-whisper (local) / ElevenLabs (opt-in)
→ detect stumbles & pauses
→ cut-plan (you approve) never auto-cuts blindly
→ cut (EDL, padded by cut type)
→ subtitles
→ motion graphics (optional) hyperframes
→ color (LUT)
→ render correctness-checked output
Status & roadmap
v0.1 ships a working talking-head pipeline behind one command. Growth (shorts, podcast, B-roll, diarization, denoise, background replacement) lands as plugins.
Credits & license
Built on the shoulders of open-source tools — see CREDITS.md. Architecture ideas (tool registry, checkpoints) inspired by OpenMontage; no code copied.
MIT © 2026 ArtCog. The bundled engines keep their own licenses.