georgismitev/mcp-salesforce-server

MCP Salesforce Connector

A Model Context Protocol (MCP) server implementation for Salesforce integration, allowing LLMs to interact with Salesforce data through SOQL queries and SOSL searches. This project provides Docker containerization for easy deployment.

Recent Updates

• New Streaming Server: Added SSE-based streaming MCP server as the main entry point for Docker workflow
• Health Check Endpoints: Added /health and /metrics endpoints for Docker health checks
• Debugging Support: Added debug mode for troubleshooting (--debug flag)
• Clean Docker Workflow: Added scripts for clean Docker management
• Improved Error Handling: Better error handling and logging for MCP server communication
• Multiple Deployment Options: Support for both Cloud Run (serverless) and VM deployment methods

Features

• Execute SOQL (Salesforce Object Query Language) queries
• Perform SOSL (Salesforce Object Search Language) searches
• Retrieve metadata for Salesforce objects, including field names, labels, and types
• Retrieve, create, update, and delete records
• Execute Tooling API requests
• Execute Apex REST requests
• Make direct REST API calls to Salesforce
• SSE-based streaming server for real-time communication
• Health check endpoints for container monitoring

Deployment Options

This server can be deployed in two ways:

1. Google Cloud Run (Serverless): For a fully managed serverless experience.

   • See for detailed instructions.

2. Google Compute Engine VM: For more consistent performance and easier troubleshooting.

   • See for detailed instructions.

Note: There are some known issues with Salesforce authentication in Cloud Run environments. If you encounter authentication problems with Cloud Run, the VM deployment option is recommended as a workaround.

Docker Setup

This project includes scripts for running the MCP server in a Docker container:

Using Docker Script

1. Start the server:

   ./scripts/run_mcp_server.sh
   

   Additional options:

   ./scripts/run_mcp_server.sh --debug --port 8080 --log-level debug
   

Using Streaming Server Locally

1. Start the streaming server:

   ./scripts/run_mcp_server.sh --port 8080
   

   Access the SSE endpoint at http://localhost:8080/sse

2. Run in debug mode (if needed):

   ./scripts/run_mcp_server.sh --debug
   

3. Verify the server is working:

   ./scripts/verify_mcp_server.sh
   

4. Restart the server if needed:

   ./scripts/restart_mcp_server.sh
   

Configuration

Salesforce Authentication Methods

This server supports two authentication methods:

• OAuth (Recommended): Set SALESFORCE_ACCESS_TOKEN and SALESFORCE_INSTANCE_URL as environment variables.
• Username/Password (Legacy): If SALESFORCE_ACCESS_TOKEN and SALESFORCE_INSTANCE_URL are not set, the server will fall back to using SALESFORCE_USERNAME, SALESFORCE_PASSWORD, and SALESFORCE_SECURITY_TOKEN.

Streaming Server Architecture

This project uses an SSE (Server-Sent Events) based streaming server as the main entry point for Docker deployments. The streaming server offers several advantages over the traditional stdio-based MCP server:

• Web-based communication: Uses HTTP and SSE for communication instead of stdio, making it more suitable for web-based clients and Docker environments
• Health monitoring: Provides /health and /metrics endpoints for monitoring the server's status
• Persistent connections: Maintains persistent connections with clients, allowing for real-time updates
• Docker-friendly: Designed to work seamlessly in containerized environments

Endpoints

• /sse: The main SSE endpoint for MCP protocol communication
• /health: Health check endpoint that returns status 200 if the server is running properly
• /metrics: Provides additional metrics about the server's state

Environment Variables

The streaming server uses the following environment variables:

• SALESFORCE_USERNAME, SALESFORCE_PASSWORD, SALESFORCE_SECURITY_TOKEN: Salesforce credentials (username-based authentication)
• SALESFORCE_ACCESS_TOKEN, SALESFORCE_INSTANCE_URL: Salesforce credentials (token-based authentication)
• PORT: The port the server listens on (default: 8080)
• LOG_LEVEL: Set logging level (default: info, options: debug, info, warning, error)
• MCP_DEBUG: Enable debug logging (default: false)

Troubleshooting

If you encounter issues with the MCP server, refer to the file for common issues and solutions.

Clean Start

To completely clean up all Docker resources related to this project and start fresh:

# Clean up all Docker containers and images
./scripts/cleanup_docker.sh

# Rebuild and start the container
./scripts/run_mcp_server.sh

This will remove all containers and images related to the MCP server and give you a clean starting point.