signal-messenger-standalone_skill

Overview

This skill provides a complete Signal messenger integration for OpenClaw agents, enabling secure text and voice messaging with role-based contact controls. It supports automatic voice transcription, TTS voice replies, conversation history, new contact triage, typing indicators, read receipts, and instant wake-on-message via the OpenClaw hooks API. Designed for secure, auditable messaging with owner/trusted/untrusted permissions.

How this skill works

A cron-backed poll script uses signal-cli to receive messages, logs them per contact, flags new contacts for triage, and posts wakes to OpenClaw so the agent can handle messages immediately. Incoming voice attachments are converted and transcribed (Whisper or other STT). Replies are generated by the agent, optionally synthesized to audio, converted to m4a, and sent via signal-cli with typing indicators and read receipts.

When to use it

• Setting up Signal messaging for an OpenClaw agent with secure contact management.
• Automatically transcribing incoming voice notes for processing or logging.
• Sending spoken replies to users via TTS and audio attachments.
• Enforcing owner/trusted/untrusted permissions to reduce prompt-injection risk.
• Triggering immediate agent wake-ups on new Signal messages.

Best practices

• Register a dedicated phone number and keep signal-cli and Java up to date.
• Maintain an explicit ALLOWLIST for trusted contacts and triage unknown senders.
• Run Whisper or an STT engine locally if privacy and latency are priorities.
• Use ffmpeg to normalize audio before transcription and TTS conversion.
• Enable OpenClaw wake hooks to avoid polling delays and ensure instant handling.

Example use cases

• A field agent sends a voice note; the agent transcribes it, logs context, and replies with a synthesized audio confirmation.
• An owner receives an approval request when an unknown UUID contacts the agent and can add them to the trusted list.
• Automated alerts from the agent posted to Signal with attachments and read receipts for auditability.
• Using per-contact conversation logs to provide context-aware responses across multiple message turns.