MCP Servers

MCP (Model Context Protocol) Servers extend agent capabilities with custom tools and resources.

Looking for the Fide MCP Server?

If you want to connect Claude Desktop or Cursor TO Fide:

What are MCP Servers?

MCP is an open protocol that allows AI agents to connect to external servers that provide:

• Tools - Custom functions agents can call
• Resources - Data and context agents can access
• Prompts - Pre-built prompt templates

Why MCP?

MCP standardizes how agents interact with external systems:

• Interoperability - Works across different AI providers
• Security - Controlled access to capabilities
• Extensibility - Add new capabilities without code changes

Connecting MCP Servers

Supported Transports

• Stdio - Local process communication
• HTTP+SSE - Remote server over HTTP

Configuration

{
  "command": "npx",
  "args": ["-y", "@modelcontextprotocol/server-filesystem"],
  "env": {
    "ALLOWED_DIRECTORIES": "/path/to/data"
  }
}

Available MCP Servers

Official Servers

• Filesystem - File read/write access
• GitHub - Repository operations
• Postgres - Database queries
• Brave Search - Web search

Community Servers

The MCP ecosystem is growing with community-built servers for various services.

Security Considerations

• Allow-listing - Only approved servers can connect
• Sandboxing - Servers run in isolated environments
• Permission Scoping - Limit what tools are exposed
• Audit Logging - Track all MCP calls