MCP ExplorerExplorer

Mcp Voice Agent

@Heminoidon 16 days ago
1 MIT
FreeCommunity
AI Systems
This project implements a voice agent that combines web search capabilities via Firecrawl with Supabase database operations through MCP (Model Context Protocol).

Overview

What is Mcp Voice Agent

mcp-voice-agent is a voice agent project that integrates web search capabilities through Firecrawl with database operations using Supabase, facilitated by the Model Context Protocol (MCP).

Use cases

Use cases include creating a voice-activated customer service agent, developing an interactive educational assistant, and implementing a voice-controlled database query system.

How to use

To use mcp-voice-agent, ensure Python 3.x is installed, install the required packages with ‘pip install -r requirements.txt’, configure the environment variables in the .env file, and run the agent with ‘python agent.py’.

Key features

Key features include real-time web search via Firecrawl, integration with Supabase for database operations, and voice interaction capabilities such as Voice Activity Detection (VAD), speech-to-text using AssemblyAI, language processing with OpenAI GPT-4, and text-to-speech using OpenAI.

Where to use

mcp-voice-agent can be used in various fields including customer support, virtual assistants, educational tools, and any application requiring voice interaction and data retrieval.

Content

MCP-powered voice agent

This project implements a voice agent that combines web search capabilities via Firecrawl with Supabase database operations through MCP (Model Context Protocol).

Installation

Ensure you have Python 3.x installed and run:

pip install -r requirements.txt

Implementation: agent.py

This implementation uses AssemblyAI’s services for speech-to-text, along with Firecrawl for web search and Supabase for database operations.

Requirements

  • Firecrawl API key
  • Supabase access token
  • OpenAI API key
  • AssemblyAI API key
  • LiveKit credentials

Setup

Copy .env.example to .env and configure the following environment variables:

FIRECRAWL_API_KEY=your_firecrawl_api_key
SUPABASE_ACCESS_TOKEN=your_supabase_token
OPENAI_API_KEY=your_openai_api_key
ASSEMBLYAI_API_KEY=your_assemblyai_api_key
LIVEKIT_URL=your_livekit_url
LIVEKIT_API_KEY=your_livekit_api_key
LIVEKIT_API_SECRET=your_livekit_api_secret

Running

Start the agent using:

python agent.py

The agent will:

  1. Connect to LiveKit
  2. Initialize the MCP server for Supabase integration
  3. Set up voice interaction capabilities
  4. Start listening for user input

Features

  • Real-time web search using Firecrawl
  • Supabase database integration via MCP
  • Voice interaction capabilities:
    • Silero VAD (Voice Activity Detection)
    • AssemblyAI Speech-to-Text
    • OpenAI GPT-4 for language processing
    • OpenAI TTS for text-to-speech

Contribution

Contributions are welcome! Feel free to fork this repository and submit pull requests with your improvements.

Tools

No tools

Comments