WhatsApp Daemon

The WhatsApp Daemon mode bridges WhatsApp messages to LLM providers, enabling AI-powered conversations through WhatsApp.

Prerequisites

A WhatsApp account (personal or business)
An active OpenAlloy license
At least one configured LLM provider

Setup

1. Start the Daemon

openalloy daemon --whatsapp

On first run, a QR code is displayed in the terminal. Scan it with WhatsApp (Settings > Linked Devices > Link a Device).

2. Configure Behavior

# Set the default model for WhatsApp responses
openalloy config set whatsapp.model anthropic:claude-sonnet-4-20250514

# Set response prefix (optional)
openalloy config set whatsapp.prefix "[AI] "

# Set allowed contacts (empty = all)
openalloy config set whatsapp.allowed_contacts "+1234567890,+0987654321"

3. Run as a Service

Linux (systemd):

openalloy daemon --whatsapp --install-service
sudo systemctl enable openalloy-whatsapp
sudo systemctl start openalloy-whatsapp

macOS (launchd):

openalloy daemon --whatsapp --install-service

Configuration Options

Key	Default	Description
`whatsapp.model`	Default model	LLM model for responses
`whatsapp.prefix`	`""`	Prefix added to AI responses
`whatsapp.allowed_contacts`	`""`	Comma-separated phone numbers (empty = all)
`whatsapp.max_history`	`20`	Messages to include as context
`whatsapp.system_prompt`	`""`	System prompt for conversations
`whatsapp.media_support`	`true`	Process images and documents

Commands

WhatsApp users can send commands by prefixing messages with /:

Command	Description
`/model <name>`	Switch the LLM model
`/reset`	Clear conversation history
`/status`	Show current model and stats
`/help`	Show available commands

Logging

Daemon logs are written to ~/.config/openalloy/logs/whatsapp.log.

tail -f ~/.config/openalloy/logs/whatsapp.log

Enterprise Features

With an Enterprise license, the WhatsApp Daemon supports:

Webhook telemetry: HMAC-signed usage events pushed to your endpoint
Multi-user: Multiple WhatsApp accounts on a single license
Usage dashboard: Per-conversation and per-model breakdowns