Ephor Collaboration MCP server

The Model Context Protocol (MCP) server enables collaborative debates between multiple AI agents, allowing them to discuss and reach consensus on user prompts. It facilitates multi-turn conversations where LLMs can register as participants, submit responses, and engage in deliberative discussion until they reach agreement.

Installation

To install the MCP server, you'll need Bun.js installed on your system:

# Install dependencies
bun install

Running the Server

After installing dependencies, you can start the server:

# Build the TypeScript code
bun run build

# Start the server
bun run dev

You can also use the MCP Inspector to test and debug your server:

# Run the server with MCP Inspector
bun run inspect

Server Endpoints

The MCP server exposes two primary endpoints:

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

Available Tools

register-participant

This tool allows an LLM to join a collaboration session with its initial response.

// Example tool call
const result = await client.callTool({
  name: 'register-participant',
  arguments: {
    name: 'Socrates',
    prompt: 'What is the meaning of life?',
    initial_response: 'The meaning of life is to seek wisdom through questioning...',
    persona_metadata: {
      style: 'socratic',
      era: 'ancient greece'
    } // Optional
  }
});

The server waits for a 3-second registration period after the last participant joins before responding with all participants' initial responses.

submit-response

This tool allows an LLM to submit follow-up responses during the debate.

// Example tool call
const result = await client.callTool({
  name: 'submit-response',
  arguments: {
    sessionId: 'EPH4721R-Socrates', // Session ID received after registration
    prompt: 'What is the meaning of life?',
    response: 'In response to Plato, I would argue that...'
  }
});

get-responses

This tool retrieves all responses from the debate session.

// Example tool call
const result = await client.callTool({
  name: 'get-responses',
  arguments: {
    sessionId: 'EPH4721R-Socrates', // Session ID received after registration
    prompt: 'What is the meaning of life?' // Optional
  }
});

The response includes all participants' contributions in chronological order.

get-session-status

This tool checks if the registration waiting period has elapsed.

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

Collaborative Debate Workflow

1. LLMs register as participants with their initial responses to the prompt
2. The server waits 3 seconds after the last registration
3. When the registration period ends, all participants receive all initial responses
4. Participants can then submit follow-up responses, responding to each other
5. The debate continues until participants reach a consensus

Deployment

You can deploy the MCP server to an EC2 instance using Docker:

Prerequisites

• EC2 instance running Amazon Linux 2 or Ubuntu
• Security group allowing inbound traffic on port 62887
• SSH access to the instance

Quick Deployment

# Clone the repository
git clone <your-repository-url>
cd <repository-directory>

# Make the deployment script executable
chmod +x deploy.sh

# Run the deployment script
./deploy.sh

Manual Deployment

# Build the Docker image
docker-compose build

# Start the container
docker-compose up -d

# Verify the container is running
docker-compose ps

Once deployed, your MCP server will be accessible at:

• http://<ec2-public-ip>:62887/sse - SSE endpoint
• http://<ec2-public-ip>:62887/messages - Messages endpoint

Important: Make sure port 62887 is open in your EC2 security group!

How to install this MCP server

For Claude Code

To add this MCP server to Claude Code, run this command in your terminal:

claude mcp add-json "ephor-collaboration" '{"command":"npx","args":["-y","ephor-mcp-collaboration"]}'

See the official Claude Code MCP documentation for more details.

For Cursor

There are two ways to add an MCP server to Cursor. The most common way is to add the server globally in the ~/.cursor/mcp.json file so that it is available in all of your projects.

If you only need the server in a single project, you can add it to the project instead by creating or adding it to the .cursor/mcp.json file.

Adding an MCP server to Cursor globally

To add a global MCP server go to Cursor Settings > Tools & Integrations and click "New MCP Server".

When you click that button the ~/.cursor/mcp.json file will be opened and you can add your server like this:

{
    "mcpServers": {
        "ephor-collaboration": {
            "command": "npx",
            "args": [
                "-y",
                "ephor-mcp-collaboration"
            ]
        }
    }
}

Adding an MCP server to a project

To add an MCP server to a project you can create a new .cursor/mcp.json file or add it to the existing one. This will look exactly the same as the global MCP server example above.

How to use the MCP server

Once the server is installed, you might need to head back to Settings > MCP and click the refresh button.

The Cursor agent will then be able to see the available tools the added MCP server has available and will call them when it needs to.

You can also explicitly ask the agent to use the tool by mentioning the tool name and describing what the function does.

For Claude Desktop

To add this MCP server to Claude Desktop:

1. Find your configuration file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json
• Linux: ~/.config/Claude/claude_desktop_config.json

2. Add this to your configuration file:

{
    "mcpServers": {
        "ephor-collaboration": {
            "command": "npx",
            "args": [
                "-y",
                "ephor-mcp-collaboration"
            ]
        }
    }
}

3. Restart Claude Desktop for the changes to take effect