OpenClaw

https://openclawlab.com/en/docs
Admin
OpenClaw is an AI agent platform that connects to multiple messaging channels and enables...
Tokens:605,746
Snippets:5,246
Trust Score:4.6
Update:7 months ago
Show doc for...
Tokens:
Raw
### LiteLLM Proxy Setup

Source: https://openclawlab.com/en/docs/providers/litellm

Instructions for installing LiteLLM and starting the proxy server.

```APIDOC
## Install LiteLLM Proxy

```bash
pip install 'litellm[proxy]'
```

## Start LiteLLM Proxy

```bash
litellm --model claude-opus-4-6
```

This command starts the LiteLLM proxy, making it available for OpenClaw to connect to.
```

--------------------------------

### OpenClaw CLI Commands for Setup

Source: https://openclawlab.com/en/docs/start/openclaw

These commands are used to initiate the OpenClaw setup process, including logging in to channels and starting the gateway. Ensure OpenClaw is installed and onboarded before running these.

```bash
openclaw channels login

```

```bash
openclaw gateway --port 18789

```

```bash
openclaw setup

```

```bash
openclaw dashboard

```

--------------------------------

### Nix Quick Start Prompt

Source: https://openclawlab.com/en/docs/install/nix

A prompt to guide an AI agent in setting up nix-openclaw on a Mac, including dependency checks, flake creation, bot setup, secret management, and verification steps.

```plaintext
I want to set up nix-openclaw on my Mac.
Repository: github:openclaw/nix-openclaw

What I need you to do:
1. Check if Determinate Nix is installed (if not, install it)
2. Create a local flake at ~/code/openclaw-local using templates/agent-first/flake.nix
3. Help me create a Telegram bot (@BotFather) and get my chat ID (@userinfobot)
4. Set up secrets (bot token, Anthropic key) - plain files at ~/.secrets/ is fine
5. Fill in the template placeholders and run home-manager switch
6. Verify: launchd running, bot responds to messages

Reference the nix-openclaw README for module options.
```

--------------------------------

### Bootstrap OpenClaw Setup

Source: https://openclawlab.com/en/docs/start/setup

Initializes the OpenClaw environment. This command can be run globally or from within the project repository. It ensures the necessary configurations and workspaces are set up.

```bash
openclaw setup
```

--------------------------------

### Run Docker Setup Script for OpenClaw

Source: https://openclawlab.com/en/docs/install/docker

Executes the primary Docker setup script to build the gateway image, run the onboarding wizard, and start the gateway via Docker Compose. This is the recommended quick start for a containerized OpenClaw environment. It generates a gateway token and writes it to the .env file.

```bash
./docker-setup.sh

```

--------------------------------

### Nostr Channel Quick Setup

Source: https://openclawlab.com/en/docs/channels/nostr

A step-by-step guide to quickly set up the Nostr channel, including key generation, configuration, and Gateway restart.

```APIDOC
## Quick Setup

1.  **Generate a Nostr Keypair (if needed):**
    ```bash
    # Using nak
    nak key generate
    ```

2.  **Add to OpenClaw Configuration:**
    ```json
    {
      "channels": {
        "nostr": {
          "privateKey": "${NOSTR_PRIVATE_KEY}"
        }
      }
    }
    ```

3.  **Export the Private Key:**
    ```bash
    export NOSTR_PRIVATE_KEY="nsec1..."
    ```

4.  **Restart the Gateway.**
```

--------------------------------

### Openclaw Onboarding with OpenCode Zen

Source: https://openclawlab.com/en/docs/reference/wizard

Example of non-interactive onboarding using OpenCode Zen. This configuration sets up the gateway with the necessary API key and network parameters.

```bash
openclaw onboard --non-interactive \
  --mode local \
  --auth-choice opencode-zen \
  --opencode-zen-api-key "$OPENCODE_API_KEY" \
  --gateway-port 18789 \
  --gateway-bind loopback
```

--------------------------------

### Start OpenClaw Gateway in Watch Mode

Source: https://openclawlab.com/en/docs/start/setup

Installs project dependencies and starts the OpenClaw Gateway in watch mode using pnpm. This enables hot reloading for TypeScript changes, ideal for the bleeding-edge workflow.

```bash
pnpm install
pnpm gateway:watch
```

--------------------------------

### Install and Connect Tailscale on VPS

Source: https://openclawlab.com/en/docs/help/faq

Steps to install Tailscale on a Virtual Private Server (VPS) and connect it to your Tailscale network. This involves running a script to install Tailscale and then logging in.

```bash
curl -fsSL https://tailscale.com/install.sh | sh
sudo tailscale up
```

--------------------------------

### Basic 'openclaw onboard' Commands

Source: https://openclawlab.com/en/docs/cli/onboard

These commands initiate the interactive onboarding wizard. You can specify different flows like 'quickstart' or 'manual', or set up a remote gateway by providing its WebSocket URL.

```bash
openclaw onboard
openclaw onboard --flow quickstart
openclaw onboard --flow manual
openclaw onboard --mode remote --remote-url ws://gateway-host:18789
```

--------------------------------

### Openclaw Onboarding with Vercel AI Gateway API Key

Source: https://openclawlab.com/en/docs/reference/wizard

Example for non-interactive onboarding with Vercel AI Gateway API key. It sets up the gateway with the specified authentication and network configurations.

```bash
openclaw onboard --non-interactive \
  --mode local \
  --auth-choice ai-gateway-api-key \
  --ai-gateway-api-key "$AI_GATEWAY_API_KEY" \
  --gateway-port 18789 \
  --gateway-bind loopback
```

--------------------------------

### Openclaw Onboarding with Gemini API Key

Source: https://openclawlab.com/en/docs/reference/wizard

Example of non-interactive onboarding using a Gemini API key. This configuration sets up the gateway with specific authentication and network parameters.

```bash
openclaw onboard --non-interactive \
  --mode local \
  --auth-choice gemini-api-key \
  --gemini-api-key "$GEMINI_API_KEY" \
  --gateway-port 18789 \
  --gateway-bind loopback
```

--------------------------------

### Check OpenClaw Health Status

Source: https://openclawlab.com/en/docs/start/setup

Verifies the operational status of the OpenClaw Gateway and its connected components. This is a crucial step for sanity checking the setup and diagnosing potential issues.

```bash
openclaw health
```

--------------------------------

### Setup Anthropic Token for Models

Source: https://openclawlab.com/en/docs/cli

This snippet demonstrates how to set up authentication for the Anthropic provider using a setup token. It first shows the general setup command and then the specific command for the Anthropic provider, followed by a command to check the models' status.

```bash
claude setup-token
openclaw models auth setup-token --provider anthropic
openclaw models status
```

--------------------------------

### CLI Setup for OpenCode Zen Onboarding

Source: https://openclawlab.com/en/docs/providers/opencode

These commands demonstrate how to onboard the OpenCode Zen model provider using the Openclaw CLI. It includes both interactive and non-interactive methods for setting up the API key.

```bash
openclaw onboard --auth-choice opencode-zen
# or non-interactive
openclaw onboard --opencode-zen-api-key "$OPENCODE_API_KEY"
```

--------------------------------

### Comprehensive OpenClaw Assistant Configuration

Source: https://openclawlab.com/en/docs/start/openclaw

An example of a detailed OpenClaw configuration file tailored for an 'assistant' setup. It includes settings for logging, agent model, workspace, thinking defaults, timeouts, heartbeats, WhatsApp channel restrictions, routing, and session management.

```json
{
  "logging": { "level": "info" },
  "agent": {
    "model": "anthropic/claude-opus-4-6",
    "workspace": "~/.openclaw/workspace",
    "thinkingDefault": "high",
    "timeoutSeconds": 1800,
    "heartbeat": { "every": "0m" },
  },
  "channels": {
    "whatsapp": {
      "allowFrom": ["+15555550123"],
      "groups": {
        "*": { "requireMention": true },
      },
    },
  },
  "routing": {
    "groupChat": {
      "mentionPatterns": ["@openclaw", "openclaw"],
    },
  },
  "session": {
    "scope": "per-sender",
    "resetTriggers": ["/new", "/reset"],
    "reset": {
      "mode": "daily",
      "atHour": 4,
      "idleMinutes": 10080,
    },
  },
}

```

--------------------------------

### Restart macOS App from Source

Source: https://openclawlab.com/en/docs/start/setup

A shell script to restart the macOS application from its source code. This is part of the bleeding-edge workflow, allowing the macOS app to run with the latest changes.

```bash
./scripts/restart-mac.sh
```

--------------------------------

### Signal-cli Installation Script

Source: https://openclawlab.com/en/docs/reference/wizard

This section describes the process of installing `signal-cli` using the wizard, which involves downloading release assets, storing them in a specific directory, and updating the configuration. It notes JVM requirements and native build usage.

```bash
# The wizard downloads the appropriate release asset for signal-cli.
# It stores the executable under ~/.openclaw/tools/signal-cli/<version>/.
# Finally, it updates the config with channels.signal.cliPath.
```

--------------------------------

### Running Openclaw Container Manually (Single User)

Source: https://openclawlab.com/en/docs/install/podman

This example demonstrates how to manually run the Openclaw container for a single-user setup. It specifies user namespace options and mounts for configuration and workspace directories.

```bash
podman run --userns=keep-id -v ~/.openclaw:/root/.openclaw -v <your_config_path>:/etc/openclaw -v <your_workspace_path>:/workspace openclaw
```

--------------------------------

### Send Media Attachment (Text)

Source: https://openclawlab.com/en/docs/start/openclaw

This example demonstrates how an agent can send an outbound attachment by specifying 'MEDIA:' followed by a path or URL on its own line. OpenClaw will extract and send this as a media attachment.

```text
Here’s the screenshot.
MEDIA:https://example.com/screenshot.png
```

--------------------------------

### Quick Setup for Synthetic Provider

Source: https://openclawlab.com/en/docs/providers/synthetic

This command initiates the onboarding process for the Synthetic provider, allowing you to authenticate using an API key. It's a prerequisite for using Synthetic models with OpenClaw.

```bash
openclaw onboard --auth-choice synthetic-api-key
```

--------------------------------

### Run OpenClaw Gateway from Local Repo

Source: https://openclawlab.com/en/docs/start/setup

Executes the OpenClaw Gateway directly from the local repository after building. This is useful for development and testing purposes, allowing specific port and verbosity configurations.

```bash
node openclaw.mjs gateway --port 18789 --verbose
```

--------------------------------

### Openclaw Onboarding with Synthetic API Key

Source: https://openclawlab.com/en/docs/reference/wizard

This command demonstrates non-interactive onboarding using a Synthetic API key, configuring the gateway with the specified authentication and network settings.

```bash
openclaw onboard --non-interactive \
  --mode local \
  --auth-choice synthetic-api-key \
  --synthetic-api-key "$SYNTHETIC_API_KEY" \
  --gateway-port 18789 \
  --gateway-bind loopback
```

--------------------------------

### Install Binaries in Dockerfile (Dockerfile)

Source: https://openclawlab.com/en/docs/install/hetzner

This Dockerfile snippet demonstrates how to install external binaries like gog, goplaces, and wacli into the Docker image at build time. It uses `curl` to download archives, `tar` to extract them into `/usr/local/bin`, and `chmod` to make them executable. This ensures binaries are available when the container starts.

```dockerfile
FROM node:22-bookworm

RUN apt-get update && apt-get install -y socat && rm -rf /var/lib/apt/lists/*

# Example binary 1: Gmail CLI
RUN curl -L https://github.com/steipete/gog/releases/latest/download/gog_Linux_x86_64.tar.gz \
  | tar -xz -C /usr/local/bin && chmod +x /usr/local/bin/gog

# Example binary 2: Google Places CLI
RUN curl -L https://github.com/steipete/goplaces/releases/latest/download/goplaces_Linux_x86_64.tar.gz \
  | tar -xz -C /usr/local/bin && chmod +x /usr/local/bin/goplaces

# Example binary 3: WhatsApp CLI
RUN curl -L https://github.com/steipete/wacli/releases/latest/download/wacli_Linux_x86_64.tar.gz \
  | tar -xz -C /usr/local/bin && chmod +x /usr/local/bin/wacli

# Add more binaries below using the same pattern

WORKDIR /app
COPY package.json pnpm-lock.yaml pnpm-workspace.yaml .npmrc ./
COPY ui/package.json ./ui/package.json
COPY scripts ./scripts

RUN corepack enable
RUN pnpm install --frozen-lockfile

COPY . .
RUN pnpm build
RUN pnpm ui:install
RUN pnpm ui:build

ENV NODE_ENV=production

CMD ["node","dist/index.js"]
```

--------------------------------

### Openclaw Onboarding with Moonshot API Key

Source: https://openclawlab.com/en/docs/reference/wizard

Example of non-interactive onboarding using a Moonshot API key. This configuration sets up the gateway with the required authentication and network parameters.

```bash
openclaw onboard --non-interactive \
  --mode local \
  --auth-choice moonshot-api-key \
  --moonshot-api-key "$MOONSHOT_API_KEY" \
  --gateway-port 18789 \
  --gateway-bind loopback
```

--------------------------------

### Manual Docker Compose Flow for OpenClaw

Source: https://openclawlab.com/en/docs/install/docker

Provides the manual steps to build the OpenClaw Docker image, run the onboard command, and start the gateway using Docker Compose. This method is an alternative to the automated setup script.

```bash
docker build -t openclaw:local -f Dockerfile .
docker compose run --rm openclaw-cli onboard
docker compose up -d openclaw-gateway

```

--------------------------------

### Openclaw Lab Recommended Starter Configuration (JSON)

Source: https://openclawlab.com/en/docs/gateway/configuration-examples

A recommended starter configuration for Openclaw Lab, including identity settings, agent model, and more detailed channel configurations like requiring mentions in WhatsApp groups. This configuration expands on the minimum setup with additional customization options.

```json
{
  "identity": {
    "name": "Clawd",
    "theme": "helpful assistant",
    "emoji": "🦞",
  },
  "agent": {
    "workspace": "~/.openclaw/workspace",
    "model": { "primary": "anthropic/claude-sonnet-4-5" },
  },
  "channels": {
    "whatsapp": {
      "allowFrom": ["+15555550123"],
      "groups": { "*": { "requireMention": true } },
    },
  },
}
```

--------------------------------

### OpenClaw Per-Hook Configuration Example (JSON)

Source: https://openclawlab.com/en/docs/automation/hooks

Demonstrates how to configure individual hooks with custom settings, such as enabling a hook and defining environment variables specific to that hook. This allows for fine-grained control over hook behavior.

```json
{
  "hooks": {
    "internal": {
      "enabled": true,
      "entries": {
        "my-hook": {
          "enabled": true,
          "env": {
            "MY_CUSTOM_VAR": "value"
          }
        }
      }
    }
  }
}
```

--------------------------------

### Enable Systemd Lingering for User Service

Source: https://openclawlab.com/en/docs/start/setup

Enables the systemd linger option for the current user. This prevents user services, like the OpenClaw Gateway, from being stopped on logout or idle, ensuring continuous operation on Linux.

```bash
sudo loginctl enable-linger $USER
```

--------------------------------

### Install OpenClaw CLI (Run Onboarding)

Source: https://openclawlab.com/en/docs/install/installer

Installs the OpenClaw CLI and automatically runs the onboarding process after installation. This is convenient for initial setup and configuration.

```shell
curl -fsSL --proto '=https' --tlsv1.2 https://openclaw.ai/install-cli.sh | bash -s -- --onboard
```

--------------------------------

### Openclaw Onboarding with Z.AI API Key

Source: https://openclawlab.com/en/docs/reference/wizard

This command shows how to perform non-interactive onboarding using a Z.AI API key, configuring the gateway with the necessary authentication and network settings.

```bash
openclaw onboard --non-interactive \
  --mode local \
  --auth-choice zai-api-key \
  --zai-api-key "$ZAI_API_KEY" \
  --gateway-port 18789 \
  --gateway-bind loopback
```

--------------------------------

### OpenClaw Installation Methods

Source: https://openclawlab.com/en/docs/help/faq

Compares the 'hackable git install' and 'npm install' methods for OpenClaw, highlighting their use cases for contributors versus general users.

```APIDOC
## OpenClaw Installation Methods

### Description
This section compares two primary methods for installing OpenClaw: the 'hackable git install' and the 'npm install'. It outlines the advantages of each method, catering to different user needs, such as development and contribution versus straightforward usage.

### Method
N/A (Installation Comparison)

### Endpoint
N/A (Installation Comparison)

### Parameters
N/A

### Request Body
N/A

### Request Example
N/A

### Response
#### Success Response (200)
N/A

#### Response Example
N/A

### Comparison
- **Hackable (git) install:**
  - Provides a full source checkout, allowing for local modifications and contributions.
  - Best suited for developers and contributors who need to patch code or documentation.
  - Requires running local builds.
- **npm install:**
  - Installs OpenClaw as a global CLI tool without direct access to the repository.
  - Best suited for users who want to "just run it" without modifying the source code.
  - Updates are managed through npm dist-tags.

### Docs
- Getting started
- Updating
```

--------------------------------

### Create macOS VM with Lume

Source: https://openclawlab.com/en/docs/install/macos-vm

Creates a new macOS virtual machine named 'openclaw' using Lume. It specifies the operating system as 'macos' and fetches the latest available IPSW file for installation. A VNC window will open to guide through the initial setup.

```bash
lume create openclaw --os macos --ipsw latest
```

--------------------------------

### Remove OpenClaw CLI Installation

Source: https://openclawlab.com/en/docs/install/uninstall

Commands to remove the OpenClaw CLI installed via npm, pnpm, or bun. Choose the command corresponding to your installation method.

```bash
npm rm -g openclaw

```

```bash
pnpm remove -g openclaw

```

```bash
bun remove -g openclaw

```

--------------------------------

### Anthropic Setup-Token Live Test Setup

Source: https://openclawlab.com/en/docs/help/testing

Instructions for setting up and running live tests to verify Anthropic's Claude Code CLI setup-token or a pasted setup-token profile. This includes enabling the test and specifying the token source, either via a profile or a raw token value.

```bash
export OPENCLAW_LIVE_SETUP_TOKEN=1
export OPENCLAW_LIVE_SETUP_TOKEN_PROFILE=anthropic:setup-token-test
export OPENCLAW_LIVE_SETUP_TOKEN_VALUE=sk-ant-oat01-...
export OPENCLAW_LIVE_SETUP_TOKEN_MODEL=anthropic/claude-opus-4-6
```

--------------------------------

### Run OpenClaw Onboarding Wizard

Source: https://openclawlab.com/en/docs/platforms/digitalocean

Initiates the OpenClaw onboarding process, which includes setting up model authentication, configuring channels (like Telegram, WhatsApp, Discord), generating a gateway token, and installing the gateway as a systemd service.

```bash
openclaw onboard --install-daemon
```

--------------------------------

### Remove OpenClaw macOS Application

Source: https://openclawlab.com/en/docs/install/uninstall

Deletes the OpenClaw macOS application if it was installed separately.

```bash
rm -rf /Applications/OpenClaw.app

```

--------------------------------

### Skill Installation Command

Source: https://openclawlab.com/en/docs/tools/clawhub

Installs a skill from the registry. Allows specifying a version and forcing an overwrite if the directory exists.

```bash
clawhub install <slug> [--version <version>] [--force]
```

--------------------------------

### Reset OpenClaw Installation

Source: https://openclawlab.com/en/docs/help/faq

Commands to perform a full reset of the OpenClaw installation, including non-interactive options and re-running the onboarding process. This is useful for troubleshooting or starting fresh.

```bash
openclaw reset

```

```bash
openclaw reset --scope full --yes --non-interactive

```

```bash
openclaw onboard --install-daemon

```

```bash
openclaw gateway --dev --reset

```

--------------------------------

### Install Node.js 22 (ARM64) on Raspberry Pi

Source: https://openclawlab.com/en/docs/platforms/raspberry-pi

Install Node.js version 22.x on your Raspberry Pi using the NodeSource repository. This command fetches the setup script, executes it, and then installs the Node.js package. Verify the installation by checking the Node.js and npm versions.

```bash
# Install Node.js via NodeSource
curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -
sudo apt install -y nodejs

# Verify
node --version  # Should show v22.x.x
npm --version
```

--------------------------------

### Install Project Dependencies with pnpm

Source: https://openclawlab.com/en/docs/platforms/mac/dev-setup

Installs all project-wide dependencies required for building and running the OpenClaw application. This command should be run in the root directory of the project.

```shell
pnpm install
```

--------------------------------

### Windows: Verbose Installer Output with PowerShell Tracing

Source: https://openclawlab.com/en/docs/install/installer

This script demonstrates how to enable PowerShell tracing to get verbose output from the install.ps1 script for debugging purposes. It sets the trace level, executes the installer script, and then resets the trace level.

```powershell
Set-PSDebug -Trace 1
& ([scriptblock]::Create((iwr -useb https://openclaw.ai/install.ps1))) -NoOnboard
Set-PSDebug -Trace 0
```

--------------------------------

### OpenClaw Plugin Configuration Example

Source: https://openclawlab.com/en/docs/plugins/zalouser

Configuration snippet for enabling and setting up a plugin. This example shows how to enable a plugin and configure its specific settings.

```yaml
plugins:
  entries:
    voice-call:
      enabled: true
      config:
        # Voice Call specific configuration
        some_setting: "value"
```

--------------------------------

### OpenClaw CLI Examples for Onboarding and Model Management

Source: https://openclawlab.com/en/docs/concepts/model-providers

Examples of using the OpenClaw CLI to onboard a provider with an API key, set a specific model, and list available models.

```bash
openclaw onboard --auth-choice opencode-zen
openclaw models set opencode/claude-opus-4-6
openclaw models list
```

--------------------------------

### Install OpenClaw and Node.js on Ubuntu

Source: https://openclawlab.com/en/docs/platforms/digitalocean

Installs Node.js version 22 and the OpenClaw application on an Ubuntu system. It first updates the system's package list and upgrades existing packages, then fetches and executes the NodeSource setup script for Node.js 22, installs Node.js, and finally downloads and runs the OpenClaw installation script. The installation concludes with a version check.

```bash
# Update system
apt update && apt upgrade -y

# Install Node.js 22
curl -fsSL https://deb.nodesource.com/setup_22.x | bash -
apt install -y nodejs

# Install OpenClaw
curl -fsSL https://openclaw.ai/install.sh | bash

# Verify
openclaw --version
```
Tokens:
Raw