Basic Memory

Description
Basic Memory is a local-first knowledge management MCP (Model Context Protocol) server that builds a semantic graph from your Markdown files to provide persistent memory across LLM conversations. It exposes local notes and documents as a structured, queryable memory layer for AI tools and agents.

Website
https://github.com/basicmachines-co/basic-memory

Category
AI Integration – MCP Servers

Tags

• memory
• knowledge-management
• local-files

---

Features

• Local-first knowledge layer

  • Operates on your local filesystem rather than relying on remote storage.
  • Uses your existing Markdown notes and documents as the primary data source.

• Semantic graph from Markdown

  • Parses Markdown files to build a semantic graph of concepts, entities, and relationships.
  • Enables structured navigation and retrieval of related information across documents.

• Persistent memory for LLM conversations

  • Maintains stateful memory across multiple AI/LLM sessions.
  • Allows AI agents to recall prior context without users repeatedly re-explaining projects or documents.

• Model Context Protocol (MCP) server

  • Implements the Model Context Protocol to make local knowledge accessible to compatible AI tools and agents.
  • Exposes notes and documents as structured, queryable resources through MCP tools and APIs.

• Structured, queryable memory layer

  • Presents your documents as a unified memory layer that can be queried semantically.
  • Supports AI assistants and agents in retrieving relevant facts, notes, and context from your local files.

• Markdown-centric workflow

  • Designed around Markdown files as the primary content format.
  • Integrates easily into existing note-taking and documentation workflows that already use Markdown.

• Open-source licensing

  • Released under the AGPL-3.0 license, allowing inspection, modification, and self-hosting subject to copyleft terms.

---

Pricing

• Not specified in the provided content (open-source project under AGPL-3.0 license).