Skills for Real Engineers. Straight from my .agents directory.
comfyui-core
Agent skill fromartokun/comfyui-mcp
Core ComfyUI knowledge covering workflow format, node types, pipeline patterns, and MCP tool usage
How to install comfyui-core
npx skills add artokun/comfyui-mcp# or add only this skill to Claude Code: copy the folder
plugin/skills/comfyui-core/
# into
~/.claude/skills/comfyui-core/ (all projects)
.claude/skills/comfyui-core/ (one project)
1# ComfyUI Core Knowledge23## Workflow JSON Format (API Format)45ComfyUI workflows are JSON objects mapping string node IDs to node definitions:67```json8{9 "1": {10 "class_type": "CheckpointLoaderSimple",11 "inputs": { "ckpt_name": "sd_xl_base_1.0.safetensors" },12 "_meta": { "title": "Load Checkpoint" }13 },14 "2": {15 "class_type": "CLIPTextEncode",16 "inputs": { "text": "a cat", "clip": ["1", 1] },17 "_meta": { "title": "Positive Prompt" }18 }19}20```2122### Key Rules2324- Node IDs are strings of integers (`"1"`, `"2"`, etc.)25- `class_type` is the exact Python class name of the node26- `inputs` contains both widget values (scalars) and connections (arrays)27- Connections use the format `["sourceNodeId", outputIndex]`, a 2-element array where:28 - the first element is the string node ID of the source node29 - the second element is the integer index into the source node's `output` list (0-based)30- `_meta` is optional and used for display titles only3132### Connection Examples3334```json35"model": ["1", 0] // Connect to node 1's first output (MODEL)36"clip": ["1", 1] // Connect to node 1's second output (CLIP)37"vae": ["1", 2] // Connect to node 1's third output (VAE)38"positive": ["2", 0] // Connect to node 2's first output (CONDITIONING)39"samples": ["5", 0] // Connect to node 5's first output (LATENT)40"images": ["6", 0] // Connect to node 6's first output (IMAGE)41```4243### Important: API Format vs Web UI Format44…
Excerpt shown under the repository's MIT license.
More skills in comfyui-mcp
Similar skill repositories
An agentic skills framework & software development methodology that works.
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
Public repository for Agent Skills
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More
