Documentation

Everything you need to install, configure, and run SquidClaw.

Installation

SquidClaw runs on Linux, macOS, and Windows. Choose your platform:

curl -fsSL https://squidclaw.dev/install.sh | bashbash

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

# Install SquidClaw
npm install -g squidclaw

# Run
squidclawbash

curl -fsSL https://squidclaw.dev/install.sh | bashterminal

# Install Node.js
brew install node@22

# Install SquidClaw
npm install -g squidclaw

# Run
squidclawterminal

iwr -useb https://squidclaw.dev/install.ps1 | iexPowerShell

# Install Node.js 22+ from https://nodejs.org
# Then in PowerShell:

npm install -g squidclaw

# Run
squidclawPowerShell

# Install globally
npm install -g squidclaw

# Or run without installing
npx squidclawany platform

Quick Start

npx squidclaw

# Set via environment variable
export ANTHROPIC_API_KEY=sk-ant-...
# or
export OPENAI_API_KEY=sk-...
# or
export GOOGLE_API_KEY=AI...

# Telegram
squidclaw channels add telegram

# WhatsApp (scan QR code)
squidclaw channels login whatsapp

# Discord
squidclaw channels add discord

squidclaw

Environment Variables

CLI Commands

All commands follow the pattern: squidclaw <command> [subcommand] [options]

🌐 Gateway

🔧 Setup & Config

💬 Channels

🤖 Agents & Sessions

🧠 Models

⏰ Automation

🔒 Security & System

🌐 Advanced

Global Flags

Config File

Config lives at ~/.squidclaw/squidclaw.json. Example:

{
  "agents": {
    "defaults": {
      "workspace": "~/.squidclaw/workspace",
      "model": {
        "primary": "anthropic/claude-sonnet-4-20250514"
      }
    }
  },
  "channels": {
    "telegram": {
      "botToken": "123456:ABC...",
      "allowFrom": ["your-telegram-id"]
    },
    "whatsapp": {
      "enabled": true
    }
  },
  "gateway": {
    "mode": "local",
    "port": 18789
  }
}squidclaw.json

TTS & Voice

SquidClaw has built-in text-to-speech support. Configure it in your squidclaw.json:

Available Providers

Config Example

{
  "messages": {
    "tts": {
      "provider": "elevenlabs",
      "apiKey": "sk_...",
      "voiceId": "JBFqnCBsd6RMkjVDRZzb",
      "modelId": "eleven_multilingual_v2",
      "stability": 0.65,
      "similarityBoost": 0.85
    }
  }
}squidclaw.json

Transcription

Built-in audio transcription — send voice notes and SquidClaw converts them to text automatically.

Available Providers

Config Example

{
  "tools": {
    "media": {
      "audio": {
        "models": [
          {
            "type": "groq",
            "key": "gsk_...",
            "model": "whisper-large-v3-turbo"
          }
        ]
      }
    }
  }
}squidclaw.json

Workspace & Persona

Fresh installs automatically scaffold a workspace with personality files your agent reads every session:

Edit SOUL.md to change your agent's name, tone, and behavior. These files persist across sessions — your agent wakes up fresh but remembers everything written to these files.

tools.profile

Controls what tools your agent has access to. Default since v3.0.17 is "coding" (file read/write included).

{
  "tools": {
    "profile": "coding"
  }
}squidclaw.json

Streaming

SquidClaw can stream responses token-by-token to supported channels (Telegram, Discord). Default since v3.0.19 is "off" for clean message delivery.

{
  "channels": {
    "telegram": {
      "streaming": "off"
    }
  }
}squidclaw.json