githubJanuary 7, 2026

CallMe

A Claude Code plugin that phones you when the AI finishes a task, gets stuck, or needs a decision—enabling true asynchronous collaboration without constant monitoring.

Overview

CallMe bridges the gap between AI autonomy and human oversight through phone calls. Start a task, walk away, and receive a call when Claude needs interaction. The plugin enables multi-turn voice conversations, so you can discuss blockers or provide guidance without returning to your terminal.

The core insight: developers shouldn't babysit AI agents. CallMe treats phone calls as a communication channel, letting Claude reach out when it matters rather than requiring constant attention.

Key Features

  • Asynchronous workflow - Start tasks and get notified via phone when Claude needs input
  • Multi-turn voice conversations - Natural dialogue during calls, not just notifications
  • Device flexibility - Works with smartphones, smartwatches, and landlines
  • Composable tools - Claude can execute web searches or other actions while maintaining call context

Code Snippets

Installation

/plugin marketplace add ZeframLou/call-me
/plugin install callme@callme

Available Tools

The plugin exposes four tools to Claude:

// Begin a new call with opening message
initiate_call({ message: "I've finished the refactoring. Want me to summarize the changes?" })

// Capture user response and continue conversation
continue_call()

// Speak without waiting for response (useful before long operations)
speak_to_user({ message: "Starting the build now, this might take a few minutes..." })

// End the call session
end_call()

Environment Configuration

# Phone provider (Telnyx or Twilio)
TELNYX_API_KEY=...
TELNYX_PHONE_NUMBER=+1...
USER_PHONE_NUMBER=+1...

# Speech processing
OPENAI_API_KEY=...

# Webhook tunnel
NGROK_AUTH_TOKEN=...

Technical Details

The architecture chains four components: Claude Code plugin → local MCP server → ngrok tunnel → phone provider (Telnyx/Twilio). Audio flows through OpenAI's Whisper for speech-to-text and their TTS API for responses.

Cost runs approximately $0.03-0.04 per minute of conversation, combining phone charges ($0.007-0.014/min), speech recognition ($0.006/min), and text-to-speech (~$0.02/min).

Connections

Connections (15)