Basic Memory enables persistent knowledge building through natural language conversations with LLMs, storing information as local Markdown files. It utilizes the Model Context Protocol for LLM interaction with the knowledge base.
Basic Memory enables persistent knowledge building through natural language conversations with LLMs, storing everything as Markdown files locally. It leverages the Model Context Protocol (MCP) for LLM integration.
Key features include continuous conversation context loading, real-time local note saving, and no specialized prompting. Installation is streamlined via uv
or Smithery, which automates Claude Desktop configuration.
Basic Memory addresses the limitations of ephemeral LLM interactions by offering local-first knowledge storage, bidirectional reading/writing for both humans and LLMs, structured Markdown formatting, and a traversable knowledge graph. It uses a SQLite database for indexing.
Users can create structured notes during conversations, navigate knowledge semantically, and maintain control over their knowledge base. The system stores data in Markdown files, extracts semantic meaning, and maintains a local knowledge graph. It is designed for use with Claude Desktop, offering tools for note creation, reading, knowledge graph navigation, and searching.
basicmachines-co/basic-memory
December 2, 2024
March 28, 2025
Python