Ephor MCP

🚀 LLM Response MCP Server

An MCP (Model Context Protocol) server that allows multiple AI agents to share and read each other's responses to the same prompt.

🚀 Quick Start

This project implements an MCP server with two main tool calls:

1. submit-response: Allows an LLM to submit its response to a certain prompt.
2. get-responses: Allows an LLM to retrieve all responses from other LLMs to a specific prompt.

This enables users to ask the same question to multiple AI agents and allows the agents to read and reflect on other LLMs' answers to the question through these tools.

📦 Installation

# Install dependencies
bun install

💻 Usage Examples

Development

# Build TypeScript code
bun run build

# Start the server in development mode
bun run dev

Testing with MCP Inspector

This project supports the MCP Inspector, a tool for testing and debugging MCP servers.

# Run the server using the inspector
bun run inspect

The inspect script uses npx to run the MCP Inspector, launching a web interface in the browser to interact with your MCP server.

You'll be able to:

• Explore available tools and resources.
• Test tool calls with different parameters.
• View the server's responses.
• Debug your MCP server implementation.

Usage

The server exposes two endpoints:

• /sse - Server-Sent Events endpoint for MCP clients to connect.
• /messages - HTTP endpoint for MCP clients to send messages.

MCP Tools

submit-response

Submit an LLM's response to a prompt:

// Example tool call
const result = await client.callTool({
  name: 'submit-response',
  arguments: {
    llmId: 'claude-3-opus',  // Unique identifier of the LLM
    prompt: 'What is the meaning of life?',  // Prompt content
    response: 'The meaning of life is...'  // LLM's response
  }
});

get-responses

Retrieve all LLM responses, optionally filtering by a specific prompt:

// Example tool call
const result = await client.callTool({
  name: 'get-responses',
  arguments: {
    prompt: 'What is the meaning of life?' // Optional prompt content
  }
});

📄 License

This project is under the MIT License.

Deployment to EC2

This project provides Docker configuration for easy deployment to EC2 or any other server environment.

Prerequisites

• Run Amazon Linux 2 or Ubuntu on the EC2 instance.
• Configure the security group to allow inbound traffic on port 62886.
• Have SSH access to the instance.

Deployment Steps

1. Clone the repository to your EC2 instance:

   git clone <your-repository-url>
   cd <repository-directory>
   

2. Make the deployment script executable:

   chmod +x deploy.sh
   

3. Run the deployment script:

   ./deploy.sh
   

This script will:

• Install Docker and Docker Compose if not already installed.
• Build the Docker image.
• Start the container in detached mode.
• Display the public URL where your MCP server can be accessed.

Manual Deployment

If you prefer manual deployment:

1. Build the Docker image:

   docker-compose build
   

2. Start the container:

   docker-compose up -d
   

3. Verify the container status:

   docker-compose ps
   

Accessing the Server

After deployment, your MCP server will be accessible at the following addresses:

• http://<ec2-public-ip>:62886/sse - SSE endpoint
• http://<ec2-public-ip>:62886/messages - Messages endpoint

Ensure that the security group is configured correctly to allow traffic from the required networks to these ports.