Model Context Protocol (MCP)
Definition
An open protocol developed by Anthropic that standardizes how AI models connect to external data sources and tools. MCP provides a universal interface for models to access databases, APIs, file systems, and other tools through a consistent protocol — replacing custom integrations with a standard. MCP servers expose capabilities; MCP clients (AI models) consume them.
Why It Matters
MCP makes it easy for AI agents to connect to any tool — which also makes it easy for agents to execute unvalidated actions across many systems simultaneously. More tool access = more attack surface. MCP standardizes the connection, but it does not standardize the governance of those connections.
How Exogram Addresses This
Exogram ships a native MCP server that integrates directly with Claude Desktop and any MCP-compatible client. Every tool call that passes through the MCP server is evaluated by the deterministic policy engine before execution. The protocol provides the connection. Exogram provides the governance.
Related Terms
Key Takeaways
- → MCP standardizes the connection, not the governance
- → More tool access = more attack surface
- → Exogram's native MCP server adds governance to every MCP tool call
- → Works with Claude Desktop and any MCP-compatible client