mirror of
https://github.com/rookiestar28/ComfyUI-OpenClaw.git
synced 2026-08-14 08:52:45 +00:00
2.3 KiB
2.3 KiB
Deployment Recipe 1: Local-only (Default)
This is the safest and recommended configuration for most users. ComfyUI and OpenClaw run on your local machine, and no ports are exposed to the network.
Architecture
graph LR
User[Your Browser] -->|localhost:8188| ComfyUI
User -->|localhost:8188| OpenClaw[OpenClaw Extension]
OpenClaw -->|localhost:11434| Ollama[Local LLM]
OpenClaw -->|https| Providers[Cloud LLM APIs]
Configuration
1. Bind Address
Ensure ComfyUI is bound to 127.0.0.1 (loopback), not 0.0.0.0.
This is the default for ComfyUI.
Verification: Run ComfyUI and check the console output:
Starting server
To see the GUI go to: http://127.0.0.1:8188
2. OpenClaw Settings
No special configuration is required.
- Admin Token: Not required for loopback-only operations (unless
OPENCLAW_ADMIN_TOKENis explicitly set). - Webhooks: Disabled by default.
- Local LLM (optional):
- Ollama:
http://127.0.0.1:11434/v1 - LM Studio:
http://localhost:1234/v1 - Keep SSRF relax flags disabled:
OPENCLAW_ALLOW_ANY_PUBLIC_LLM_HOST=0OPENCLAW_ALLOW_INSECURE_BASE_URL=0
- Ollama:
- No-origin convenience override (optional, local-only):
- default/unset keeps strict no-origin denial
- set
OPENCLAW_LOCALHOST_ALLOW_NO_ORIGIN=trueonly when local CLI/tooling compatibility requires it - keep it unset/disabled for normal browser-only local use
- Optional log hygiene:
OPENCLAW_LOG_TRUNCATE_ON_START=1clears staleopenclaw.logcontent once at startup.
3. "Red Lines" (What NOT to do)
- ❌ Do not run with
--listen 0.0.0.0or--listen. - ❌ Do not port-forward port 8188 on your router.
- ❌ Do not leave
OPENCLAW_LOCALHOST_ALLOW_NO_ORIGIN=trueenabled longer than needed.
Testing
- Open
http://127.0.0.1:8188in your browser. - Open the OpenClaw tab in the sidebar.
- Go to Settings -> Health.
- If using Ollama, verify the daemon is reachable first via the native Ollama health/list surface
http://127.0.0.1:11434/api/tags. - In Settings -> LLM, set provider to
Ollama (Local), leave Base URL empty to use the built-inhttp://127.0.0.1:11434/v1default (or set that exact loopback URL explicitly), and click Load Models. - All checks should be green.