Memory MCP Server

Example Use Cases

• 1

  Persistent storage and retrieval of conversation memories (save, add, get, clear) across sessions.

• 2

  Automatic context window management: archive early messages when the window fills and retrieve relevant archived content when needed.

• 3

  Score relevance and create linked summaries of archived conversation segments for long-running dialogs.

Description

Memory MCP provides persistent MongoDB-backed storage for conversation memories and an orchestration layer for automatic context window caching. It archives, scores, retrieves, and summarizes conversation context to keep LLM contexts concise and relevant. The tool supports tag-based search, summary linking, and simple relevance scoring to help agents restore or condense past conversation state.