Say MCP Server

A Model Context Protocol (MCP) server that provides real-time voice notifications.

Features

• Voice Notifications: Get instant audible updates on task progress and results.

• High-Quality Voice: Utilizes a high-quality voice engine for clear and natural-sounding speech, especially for Japanese.

• Automatic Fallback: Seamlessly switches to a standard system voice if the high-quality engine is unavailable, ensuring you never miss a notification.

• Easy Integration: Designed for easy integration with AI agent environments that support the Model Context Protocol.

Installation

1. Clone the repository:

   git clone <repository-url> cd say-mcp

2. Install dependencies:

   uv sync

3. Test the installation:

   uv run python test_server.py

Usage

With an AI Agent (MCP Integration)

Add this configuration to your agent's MCP settings:

Available Tools

The server provides the following tools for your AI agent:

1. speak: Notifies you by voice.

   • Purpose: Converts text to speech for real-time, audible feedback.

   • Example:

     { "text": "The analysis is complete and the report is ready.", "speaker": 1, "speed_scale": 1.2 }

   • Parameters:

     • text (required): The message to be spoken.

     • speaker (optional): The desired speaker ID for the high-quality voice.

     • speed_scale (optional): Adjusts the speech speed for the high-quality voice.

2. check_voice_status: Checks the quality of the available voice.

   • Purpose: Confirms whether the high-quality voice engine is active or if the system will use the standard fallback voice.

High-Quality Voice Setup (Optional)

For the best audio experience, you can install a compatible high-quality voice engine (e.g., VOICEVOX). The server will automatically detect and use it when it's running.

Requirements

• Python 3.13+

• A compatible operating system (e.g., macOS for the fallback voice).

• A high-quality voice engine like VOICEVOX (optional).

License

This project is open source and available under the MIT License.