Search through notes in the knowledge base using semantic or keyword search
Create a new note in the knowledge base
Update an existing note in the knowledge base
Retrieve the content of a specific note by title
Get a list of recently created or modified notes
Basic Memory is a powerful knowledge management system that allows you to create and maintain a persistent semantic graph from your conversations with AI assistants. It stores information in standard Markdown files on your computer, making it easy to integrate with tools like Obsidian.md. The system helps you build a personal knowledge base that grows with each interaction, allowing AI assistants to reference and build upon previous conversations. By organizing information semantically, Basic Memory makes it easier to retrieve relevant information and maintain context across multiple sessions.
Basic Memory creates a persistent knowledge graph from your AI conversations, storing everything in standard Markdown files that integrate seamlessly with tools like Obsidian.md.
There are several ways to install Basic Memory:
pip install basic-memory
For Unix-based systems (macOS, Linux):
curl -sSL https://raw.githubusercontent.com/basicmachines-co/basic-memory/main/scripts/install.sh | bash
git clone https://github.com/basicmachines-co/basic-memory.git
cd basic-memory
pip install -e .
After installation, you need to configure Basic Memory:
basic-memory mcp --notes-dir /path/to/your/notes
You can also specify a specific project within your notes:
basic-memory mcp --notes-dir /path/to/your/notes --project "Project Name"
Once the MCP server is running, you can connect to it from AI assistants that support the Model Context Protocol. Basic Memory provides several tools that allow the AI to:
The system automatically creates a semantic graph from your conversations, linking related concepts and making it easy to retrieve relevant information later.
Basic Memory works particularly well with Obsidian.md:
For more detailed information, visit the Basic Memory website or refer to the GitHub repository.