learn

What is MCP (Model Context Protocol)? And how dig uses it

MCP is the open protocol that lets AI agents talk to tools and data over a standard interface. dig ships an MCP server so any compatible agent can drive a knowledge base.

MCP — the Model Context Protocol — is an open standard for connecting AI agents to tools and data through one consistent interface. Instead of a bespoke integration per app, a client (the agent) talks to any number of MCP servers (the tools), and each server advertises the tools it offers.

Why it matters

Before MCP, every agent–tool pairing was custom glue. With MCP, a tool implements the protocol once and every MCP-capable client can use it — Claude Code, Cursor, Codex, and more. One server, the whole ecosystem.

dig as an MCP server

dig ships a stdio MCP server. Register it and your agent gets the full dig surface:

{ "mcpServers": { "dig": { "command": "dig", "args": ["mcp"] } } }

The tools it exposes:

  • Read — dig_find (ranked search), dig_recall (budgeted memory), dig_drift, dig_log, dig_export. These never change state.
  • Capture — dig_retain writes a note, document, or session into memory.
  • Mutate — dig_org, dig_reconcile (preview unless you pass apply), and dig_undo. Every change is reversible.

Getting started

The fastest path is the Claude Code plugin, which bundles the server for you. Any other MCP client uses the config above — see all integrations.

Install dig · What is agent memory?