MCP (Model Context Protocol) is an open standard that defines how an AI agent connects to external sources like knowledge bases, APIs and files. A service publishes an MCP server, and the agent gets tools and data from it in a common format, with no custom integration for each model.
Anthropic introduced the protocol in 2024. Today it's supported by Claude, ChatGPT, Cursor and many other agents.
How MCP fits in with Pachca:
- Up-to-date API docs are available to agents through the Context7 MCP server
- To work with workspace data, agents use Pachca CLI, Agent Skills or the API
- Every way to connect an agent is covered in the AI agent docs on dev.pachca.com
For developers
Docs on dev.pachca.com →