n8n-nodes-openclaw-dqh

Community nodes for connecting n8n to OpenClaw-compatible HTTP endpoints.

Current status

This package currently includes one node:

• OpenClaw Chat — send a prompt to an OpenClaw-compatible HTTP endpoint and get text back.

Supported connection layouts

This package now makes the connection model explicit in the credential so users can choose the right setup.

1) n8n and OpenClaw on the same machine

Use this when:

• n8n and OpenClaw run on the same computer, container network, or private LAN
• your n8n process can reach http://127.0.0.1:3000 or another private address directly

Typical Base URL examples:

• http://127.0.0.1:3000
• http://openclaw:3000
• http://192.168.1.25:3000

Important:

• 127.0.0.1 only works if it is localhost from the n8n process itself
• if your n8n is hosted somewhere else, do not use localhost here

2) Hosted n8n to hosted OpenClaw

Use this when:

• your n8n runs on n8n Cloud or a VPS
• your OpenClaw-compatible API is also hosted on a VPS/domain and exposed publicly

Typical Base URL examples:

• https://openclaw.example.com
• https://ai.example.net

Checklist:

• endpoint must be reachable from the internet or at least from the hosted n8n environment
• use HTTPS
• protect the endpoint with a bearer token or another auth layer if exposed publicly

3) Hosted n8n to local OpenClaw via tunnel

Use this when:

• n8n is hosted remotely
• OpenClaw runs on a local laptop, desktop, or home server
• you expose OpenClaw through a tunnel or reverse proxy

Typical tunnel examples:

• Cloudflare Tunnel
• Tailscale Funnel
• reverse proxy on a public server

Typical Base URL example:

• https://my-openclaw-tunnel.example.com

Important:

• do not use http://127.0.0.1:3000 in this mode
• the tunnel URL must be reachable from your hosted n8n instance
• if you use self-signed TLS in a lab environment, enable Ignore SSL Issues only when you understand the risk

Included credentials

OpenClaw API

Fields:

• Connection Mode — choose the actual network layout
• Base URL — the OpenClaw URL that your n8n instance can reach
• API Token — optional bearer token
• Chat Path — defaults to /api/chat
• Ignore SSL Issues — optional, useful for self-signed local certs

The credential UI includes setup guidance for the selected connection mode. It also includes a built-in Test Connection action that sends a small POST request to the configured chat endpoint.

Included nodes

OpenClaw Chat

Inputs:

• Message
• System Prompt (optional)
• Session ID (optional)
• Model (optional)
• Temperature (optional)
• Max Tokens (optional)
• Include Raw Response

Outputs:

• response
• sessionId
• model
• raw (optional)

The node also shows a reminder that hosted n8n cannot reach localhost on a user laptop unless OpenClaw is exposed through a tunnel or hosted remotely.

Expected HTTP contract

This node expects an endpoint like:

POST {baseUrl}{chatPath}

Example request:

{
  "message": "Hello",
  "systemPrompt": "You are a helpful assistant",
  "sessionId": "optional-session-id",
  "model": "optional-model",
  "temperature": 0.7,
  "maxTokens": 1024
}

Supported response shape A:

{
  "response": "Hi there",
  "sessionId": "abc",
  "model": "openai-codex/gpt-5.4"
}

Supported response shape B:

{
  "choices": [
    {
      "message": {
        "content": "Hi there"
      }
    }
  ]
}

Also supported:

• { "text": "Hi there" }
• { "content": "Hi there" }
• OpenAI-style choices[0].message.content arrays containing text parts

If the endpoint returns JSON but none of those fields exist, the node throws a clear error instead of returning a raw JSON string silently.

Install for development

npm install
npm run build

Install in n8n

After publishing, install it as a community node in n8n or add it to your custom nodes setup.

Quick setup guide for users

Same machine / private network

1. Set Connection Mode to n8n and OpenClaw on the Same Machine
2. Enter a directly reachable Base URL such as http://127.0.0.1:3000
3. Keep Chat Path as /api/chat unless your adapter uses a different route
4. Add API Token if your endpoint requires it
5. Click Test Connection in the credential
6. Run the node

Hosted n8n to hosted OpenClaw

1. Set Connection Mode to Hosted n8n to Hosted OpenClaw
2. Enter your public HTTPS Base URL
3. Confirm the endpoint is reachable externally
4. Add API Token if required
5. Click Test Connection in the credential
6. Run the node

Hosted n8n to local OpenClaw via tunnel

1. Expose your local OpenClaw server through a tunnel or reverse proxy
2. Confirm the tunnel forwards requests to your local OpenClaw port
3. Set Connection Mode to Hosted n8n to Local OpenClaw via Tunnel
4. Paste the public tunnel URL into Base URL
5. Add API Token if required
6. Click Test Connection in the credential
7. Run the node

Publish

npm login
npm publish --access public

Notes before publishing

• Confirm the npm package name n8n-nodes-openclaw-dqh is available
• Update author, homepage, and repository in package.json if needed
• Build once locally before publishing
• If you want better marketplace trust, add a real GitHub repo and screenshots in the next revision