Model Context Protocol (MCP) Servers Guide (Coming Soon)

​

Understanding MCP Servers

MCP servers act as bridges between your chatbot and external services, allowing it to:

• Access and manipulate data from various sources
• Integrate with third-party services
• Enhance the chatbot’s functionality with real-time data

​

Server Types

​

SSE (Server-Sent Events) Servers

• Most MCP servers use SSE for real-time communication
• Require an endpoint URL from mcp.composio.dev
• Support various services like Google Sheets, Airtable, Trello, etc.

​

Command-based Local Servers

Run an MCP SSE proxy to turn your local MCP server into a remote one. This example repo shows you how to run an MCP proxy to turn a local MCP server into a remote / SSE MCP server. If you follow this route, you will need to make sure you expose to appropriate ports via ngrok or similar tools.

​

Setting Up MCP Servers

​

Getting Started

1. Navigate to your chatbot’s MCP Servers page
2. Browse available server templates
3. Click “Install” on your desired server

​

Configuration Steps

1. Get the SSE Endpoint:

   • Visit mcp.composio.dev
   • Select your desired service
   • Choose “Cursor” as your platform
   • Copy the generated SSE endpoint URL (Only need to copy the URL, not the whole command)

2. Using a custom local server:

   Run an MCP SSE proxy to turn your local MCP server into a remote one. This example repo shows you how to run an MCP proxy to turn a local MCP server into a remote / SSE MCP server. If you follow this route, you will need to make sure you expose to appropriate ports via ngrok or similar tools.

3. Install the Server:

   • Paste the SSE endpoint URL in the installation dialog
   • Configure server settings:
     • Active: Enable/disable the server

​

Authentication

For services requiring authentication (e.g., Google Sheets, Google Calendar):

1. Click “Install” on the service
2. Enter the SSE endpoint URL
3. The authentication will be automatically associated with the URL
4. Simply chat with the chatbot to authenticate, for example (“authenticate with google sheets”)
5. Visit the URL provided by the chatbot to complete the authentication process

​

Important Concepts

​

Active Servers Limit

• Maximum of 3 active MCP servers per chatbot
• This limit ensures optimal performance
• You can install more servers but only 3 can be active at once
• Toggle servers active/inactive as needed

​

Public vs Private Servers

• Public Servers: Available for use by other chatbots
• Private Servers: Only accessible by your chatbot
• Consider making servers public if you want to share functionality

​

Best Practices

1. Server Management:

   • Keep only necessary servers active
   • Regularly review and update server configurations
   • Test server connections after setup

2. Security:

   • Use authenticated endpoints for sensitive data
   • Keep authentication credentials secure
   • Review public/private settings regularly

3. Performance:

   • Monitor server response times
   • Deactivate unused servers
   • Use appropriate rate limits

​

Available MCP Servers

​

Google Sheets

• Store and retrieve data from spreadsheets
• Real-time updates and modifications
• Requires Google authentication

​

Airtable

• Connect to Airtable bases
• Access and modify records
• Real-time database operations

​

Trello

• Manage boards and cards
• Track tasks and projects
• Real-time updates

​

Google Calendar

• Manage events and schedules
• Set reminders
• Check availability

​

And More

• Additional servers are regularly added
• Check the templates page for the latest options

​

Troubleshooting

​

Common Issues

1. Connection Problems:

   • Verify the SSE endpoint URL
   • Check server status at mcp.composio.dev
   • Ensure authentication is valid

2. Authentication Errors:

   • Re-authenticate the service
   • Check permission settings
   • Verify API access

3. Performance Issues:

   • Review active server count
   • Check server logs
   • Monitor rate limits