Release Notes v0.1.0¶
Release Date: 2026-01-19
Why Should I Care About This Release?¶
Your AI coding assistant can now call you on the phone. Start a long-running task, walk away, and get a phone call when the AI is done, stuck, or needs a decision. No more staring at a terminal waiting for completion.
Key Highlights¶
-
Phone calls from your AI assistant - Claude Code, AWS Kiro CLI, and Gemini CLI can initiate real phone calls to discuss complex decisions, report completion, or ask clarifying questions
-
Premium voice quality - Natural conversations using ElevenLabs streaming TTS and Deepgram streaming STT, not robotic text-to-speech
-
Single binary deployment - 53 MB self-contained Go binary with no runtime dependencies. Copy one file and you're done
-
Provider-agnostic architecture - Built on the omnivoice abstraction layer, making it easy to swap TTS/STT/phone providers
What's New¶
MCP Voice Call Plugin¶
The core plugin provides four MCP tools:
| Tool | Purpose |
|---|---|
initiate_call |
Start a new call to the user with an initial message |
continue_call |
Continue conversation on an active call |
speak_to_user |
Speak without waiting for response (status updates) |
end_call |
End the call with optional goodbye message |
Multi-Assistant Support¶
Generate configuration files for your preferred AI coding tool:
go run ./cmd/generate-plugin claude . # Claude Code
go run ./cmd/generate-plugin kiro . # AWS Kiro CLI
go run ./cmd/generate-plugin gemini . # Gemini CLI
The agentplexus Stack¶
This release showcases the complete agentplexus voice AI architecture:
- omnivoice - Provider-agnostic interfaces for TTS, STT, Transport, CallSystem
- go-elevenlabs - ElevenLabs streaming TTS with natural voices
- omnivoice-deepgram - Deepgram streaming STT with accurate transcription
- omnivoice-twilio - Twilio transport and call system
- mcpkit - MCP server runtime with built-in ngrok integration
Use Cases¶
Ideal for:
- Reporting significant task completion
- Requesting clarification when blocked
- Discussing complex architectural decisions
- Walking through code changes verbally
- Multi-step processes needing back-and-forth
Cost Estimate¶
| Service | Cost |
|---|---|
| Twilio outbound calls | ~$0.014/min |
| ElevenLabs TTS | ~$0.03/min of speech |
| Deepgram STT | ~$0.0043/min |
| Total per minute | ~$0.05/min |
Getting Started¶
See the README for installation and configuration instructions.
Credits¶
Inspired by ZeframLou/call-me (TypeScript).