The Honcho MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Honcho account. It provides structured and secure access to your workspaces, peers, sessions, messages, and conclusions, so your agent can organize conversations, record context, search memories, retrieve peer representations, and reason about peers on your behalf.
- Workspace and conversation setup: Have your agent create or retrieve workspaces, peers, and sessions for ongoing interactions.
- Message-based memory: Let the agent record attributed messages, review session history, and retrieve a specific message when needed.
- Context search: Direct your agent to semantically search messages across workspaces, peers, or sessions, and search stored conclusions within a workspace.
- Peer understanding: Instruct your agent to retrieve a curated peer representation or reason over what Honcho has learned about that peer.
- Conclusion and session management: Have your agent store explicit conclusions, review saved conclusions, list existing sessions, or delete sessions you no longer need.