VideoAgent ComfyUI Video Production Orchestrator A session scoped AI orchestrator for Claude Code that turns Claude into a senior video production technical director. It routes natural language requests to 13 specialized skill modules covering the full pipeline: character image generation, video production, voice synthesis, LoRA training, and publishing all driven by ComfyUI. Why This Exists Producing AI generated video with ComfyUI involves juggling dozens of models, custom nodes, prompt styles, and hardware constraints. VideoAgent wraps all of that domain knowledge into a structured skill system that Claude reads on demand, so you can say things like: "Generate a photorealistic portrait of a dog using InstantID" "Make a talking head video where she says 'Welcome to my channel'" "Train a LoRA from these 20 reference images" "Check for new ComfyUI video models released this month" ...and get validated, hardware aware ComfyUI workflows without memorizing node names or VRAM budgets. How It Works [安装/使用命令见下方] Key design decisions: Decision Rationale Session scoped, not global Skills stay in this repo. Other Claude Code sessions are unaffected. CLAUDE.md is the orchestrator The only file Claude auto loads from a project root. Contains the routing table and behavioral instructions. Skills are read on demand markdown No build step, no registration. Claude reads skills/{name}/SKILL.md when the routing table says to. REST polling over WebSocket Claude Code can't hold persistent connections. Polling every 5s works fine for minute long video generation. Research is user triggered No cron jobs. A session start hook reminds you when data is stale. Prerequisites Claude Code installed and authenticated ComfyUI installed (local or remote) FFmpeg on PATH (for video assembly) PowerShell 7+ (for utility scripts) Windows (the launcher is a .bat file; WSL/Linux adaptation is straightforward) Quick Start 1. Clone and launch [安装/使用命令见下方] With options: [安装/使用命令见下方] 2. Scan your ComfyUI in