mirror of
https://github.com/HuangYuChuh/ComfyUI_Skills_OpenClaw.git
synced 2026-07-28 14:51:10 +00:00
33 lines
3.0 KiB
YAML
33 lines
3.0 KiB
YAML
- question: What is ComfyUI Skills for OpenClaw?
|
|
answer: An open-source tool that turns ComfyUI API-format workflows into callable skills for AI agents. Install the CLI with "pip install comfyui-skill-cli" and any agent that can run shell commands can generate images through ComfyUI.
|
|
|
|
- question: What problem does this project solve?
|
|
answer: It lets an agent take natural-language requests, map them to workflow parameters via the CLI, submit jobs to one or more ComfyUI servers, wait for completion, and download the generated images automatically.
|
|
|
|
- question: Which AI agents are supported?
|
|
answer: Any agent that can execute shell commands works, including OpenClaw, Claude Code, Codex, and others. The CLI outputs structured JSON that agents can parse directly.
|
|
|
|
- question: How do I install it?
|
|
answer: "Three steps: (1) clone the project into your agent's skills directory, (2) run 'pip install comfyui-skill-cli' to install the CLI, (3) run 'comfyui-skill server status' to verify. See the Getting Started guide for details."
|
|
|
|
- question: Can this project work with multiple ComfyUI servers?
|
|
answer: Yes. The CLI supports multi-server workflow namespaces with commands like "comfyui-skill server add" and "comfyui-skill server list". One setup can route jobs across different ComfyUI servers using the format server_id/workflow_id.
|
|
|
|
- question: What files are required for a workflow to work?
|
|
answer: Each workflow needs an API-format workflow.json exported from ComfyUI and a schema.json file that maps parameters for the agent. The CLI can auto-generate both with "comfyui-skill workflow import".
|
|
|
|
- question: Does the project include a web UI?
|
|
answer: Yes. An optional local web dashboard is included for uploading workflows, configuring servers, and editing schema mappings visually. The CLI covers all the same functionality for agent use.
|
|
|
|
- question: Why does the project use schema.json instead of exposing the whole graph?
|
|
answer: The schema layer gives the agent a smaller and safer contract with clear aliases, types, descriptions, and required flags, which makes workflow calls more reliable than exposing every node field directly.
|
|
|
|
- question: How do I import a new workflow?
|
|
answer: "For manual CLI imports, pass the raw workflow JSON as an absolute path, for example 'comfyui-skill workflow import /absolute/path/to/my-workflow.json'. It auto-detects the format (editor or API), converts if needed, and generates the schema under 'data/<server_id>/<workflow_id>/'. Then run 'comfyui-skill deps check' to verify dependencies."
|
|
|
|
- question: Can I check and install missing dependencies automatically?
|
|
answer: "Yes. Run 'comfyui-skill deps check local/my-workflow' to see what's missing, then 'comfyui-skill deps install local/my-workflow --all' to install missing custom nodes and models via ComfyUI Manager."
|
|
|
|
- question: Can I move my workflow mappings to another machine?
|
|
answer: "Yes. Use 'comfyui-skill config export' to create a portable bundle, then 'comfyui-skill config import' on the target machine. Supports dry-run preview before applying."
|