Model Context Protocol (MCP)

Model Context Protocol (MCP) gives AI applications a common way to talk to external tools and data sources.

Without something like MCP, every assistant ends up with its own custom integration for files, GitHub, databases, documentation, calendars, or internal APIs. MCP turns those integrations into servers that a compatible client can discover and call.

That is useful, but it also moves the trust boundary. If a model can call a tool, that tool needs sensible permissions. Read-only context is one thing. A tool that can write files, send messages, or deploy infrastructure deserves a much harder look.