OpenClaw & Monk

This repository contains Monk.io template to deploy OpenClaw AI assistant gateway.

Prerequisites

• Install Monk
• Register and Login Monk
• Add Cloud Provider (or run locally)

Clone Repository

git clone https://github.com/monk-io/monk-openclaw

Load Template

cd monk-openclaw
monk load MANIFEST

Add Required Secrets

# Generate and add gateway token
monk secrets add -g openclaw-gateway-token="$(openssl rand -hex 32)"

Deploy

Standalone Gateway

$ monk run openclaw/gateway

✔ Starting the run job: local/openclaw/gateway... DONE
✔ Preparing nodes DONE
✔ Checking/pulling images...
✔ [================================================] 100% alpine/openclaw:2026.2.1
✔ Checking/pulling images DONE
✔ Starting containers DONE
✔ Started local/openclaw/gateway
🔩 templates/local/openclaw/gateway
 └─🧊 Peer local
    └─🔩 templates/local/openclaw/gateway
       └─📦 openclaw-gateway running
          ├─🧩 alpine/openclaw:2026.2.1
          ├─🔌 open (public) TCP 0.0.0.0:18789 -> 18789
          └─🔌 open (public) TCP 0.0.0.0:18790 -> 18790

💡 You can inspect and manage your above stack with these commands:
	monk logs (-f) local/openclaw/gateway - Inspect logs
	monk shell     local/openclaw/gateway - Connect to the container's shell
	monk do        local/openclaw/gateway/action_name - Run defined action (if exists)
💡 Check monk help for more!

With Ollama (CPU)

# Load Ollama template first
monk load ollama/ollama

# Run OpenClaw + Ollama stack
monk run openclaw/stack-ollama

With Ollama (GPU)

# Load Ollama GPU template first
monk load ollama/ollama-gpu

# Run OpenClaw + Ollama GPU stack
monk run openclaw/stack-ollama-gpu

Pull Ollama Model

After deploying with Ollama, pull a model:

# Pull a fast model for CPU
monk do openclaw/ollama/pull-model model=qwen2.5:0.5b

# Or pull a larger model for GPU
monk do openclaw/ollama/pull-model model=llama3.3

# List installed models
monk do openclaw/ollama/list-models

Available Runnables

Runnable	Description
openclaw/gateway	Standalone OpenClaw gateway
openclaw/cli	OpenClaw CLI for management
openclaw/stack	Gateway process group
openclaw/stack-ollama	Gateway + Ollama (CPU)
openclaw/stack-ollama-gpu	Gateway + Ollama (GPU)
openclaw/ollama	Ollama CPU instance
openclaw/ollama-gpu	Ollama GPU instance

CLI Actions

# Run onboarding wizard
monk do openclaw/cli/onboard

# Login to WhatsApp
monk do openclaw/cli/channels-login

# Check channels status
monk do openclaw/cli/channels-status

# Health check
monk do openclaw/cli/health

Configuration Variables

variables:
  openclaw-image: "alpine/openclaw:2026.2.1"  # Docker image
  gateway-port: 18789                          # Gateway HTTP port
  bridge-port: 18790                           # Bridge port
  gateway-bind: "lan"                          # Bind mode (lan, loopback, public)

Secrets

Secret	Required	Description
openclaw-gateway-token	Yes	API token for gateway access
claude-ai-session-key	No	Claude AI session key
claude-web-session-key	No	Claude Web session key
claude-web-cookie	No	Claude Web cookie

Add secrets:

monk secrets add -g openclaw-gateway-token="your-token-here"

Changing the Default Model

The template uses qwen2.5:0.5b (CPU) or llama3.3 (GPU) by default. To use a different model:

1. Pull the new model:

monk do openclaw/ollama/pull-model model=mistral

2. Edit openclaw.yaml - find the gateway-ollama section and update:

"models": [
  { "id": "mistral", "name": "Mistral 7B" },
  ...
]

And set it as default:

"agents": {
  "defaults": {
    "model": {
      "primary": "ollama/mistral"
    }
  }
}

3. Reload and restart:

monk load MANIFEST
monk stop openclaw/stack-ollama
monk run openclaw/stack-ollama

Important: The model ID must match the name shown by monk do openclaw/ollama/list-models.

REST API

OpenClaw gateway exposes a REST API on port 18789.

Health Check

curl http://localhost:18789/health

Gateway Status

curl -H "Authorization: Bearer YOUR_TOKEN" http://localhost:18789/api/status

Persistence

Data is persisted under ${monk-volume-path}/openclaw:

• config/ - OpenClaw configuration
• workspace/ - Agent workspaces

Ollama models are persisted under ${monk-volume-path}/ollama.

Recommended Models for CPU

Model	Size	Speed	Command
qwen2.5:0.5b	398MB	⚡⚡⚡ Very fast	monk do openclaw/ollama/pull-model model=qwen2.5:0.5b
tinyllama	637MB	⚡⚡⚡ Very fast	monk do openclaw/ollama/pull-model model=tinyllama
llama3.2:1b	1.3GB	⚡⚡ Fast	monk do openclaw/ollama/pull-model model=llama3.2:1b
gemma2:2b	1.6GB	⚡⚡ Fast	monk do openclaw/ollama/pull-model model=gemma2:2b

Troubleshooting

Check logs

monk logs -f openclaw/gateway
monk logs -f openclaw/ollama

Enter container shell

monk shell openclaw/gateway
monk shell openclaw/ollama

Verify Ollama is running

curl http://localhost:11434/api/tags

Stop and Clean Up

# Stop stack
monk stop openclaw/stack-ollama

# Remove completely
monk purge openclaw/stack-ollama

Links

• OpenClaw Documentation
• OpenClaw Docker Image
• Ollama