Basic Memory: An AI conversational assistant that builds a personal knowledge graph

Core functions and solutions

Problem: Limitations of conversational AI

Traditional AI dialogue has obvious limitations:

-  The context disappears after the conversation ends

-  Users have to repeat information that has already been discussed

-  Unable to build on previous insights

-Knowledge  is scattered across different conversations, making it difficult to retrieve and connect

Solution: Knowledge persistence and intelligent association

Basic Memory provides innovative solutions:

1. Persistent knowledge storage : All conversation content is saved locally in the form of Markdown files, 100% controlled by the user

2. Knowledge graph construction : Notes are automatically connected through semantic links and observations to form a network structure 

3. Two-way flow : You can edit files directly or modify content through AI dialogue, always keeping in sync 

4. Conversation Continuity : Even after several weeks, you can continue the conversation from where you left off. 

5. Multi-project management : separate knowledge bases can be managed for different purposes (work, research, specific fields) 

6. Intelligent retrieval : AI can find and retrieve content only related to the current topic without uploading the entire knowledge base 

How it works

Basic Memory enables users to build and access a knowledge base through natural interactions:

1. Build on past knowledge

The AI ​​references previous notes in future conversations. For example, a user can say, "Let's continue our conversation about brewing coffee," and the AI ​​will retrieve relevant notes instead of loading all the content.

2. Natural Conversation

Users can have normal conversations with AI assistants such as Claude. For example, "Find information about hand-brewed coffee recipes," and the AI ​​will search the user's knowledge base and summarize the findings.

3. Preserve knowledge

Users can ask the AI ​​to create a note about a conversation. For example: "Create a note about coffee brewing methods", and the AI ​​will create a structured note with observations and relationships.

Technical Implementation

Basic Memory uses the Model Context Protocol (MCP) to connect to large language models, making it particularly suitable for use with Claude Desktop. Its underlying technology implementation includes:

-  All content is stored in Markdown files

-  Use SQLite database for searching and indexing

-  Extract semantic meaning from simple Markdown patterns

-Maintain  a local knowledge graph derived from documents

-  Provides two-way synchronization between files and knowledge graphs

-  Implement Model Context Protocol (MCP) for AI integration

-  Expose tools to let AI assistants traverse and manipulate knowledge graphs

-  Reference entities using memory:// URLs

File format and structure

Each Markdown file contains structural elements that are easy for AI to understand and process:

Pre-metadata

---title: Coffee Brewing Methodspermalink: coffee-brewing-methodstype: notetags:- Coffee- Brewing---

Observation content

Observations are about the fact that the content is about a topic, which can be added via specially formatted Markdown lists:

Observations - [Method] Pour-over provides a clearer mouthfeel and highlights subtle flavors - [Technique] Water temperature at 205°F (96°C) extracts the best compounds - [Principle] Freshly ground beans retain aroma and flavor

relation

Relationships are links to other topics, defining how entities are connected in the knowledge graph:

- relates_to [[coffee bean origin]] - requires [[correct grinding technique]] - affects [[flavor extraction]]

Seamless integration with Obsidian

Basic Memory is perfectly integrated with the Obsidian editor, providing powerful visual knowledge navigation functions:

Knowledge Graph Visualization

Obsidian's graph view reveals the semantic connections between notes and displays the entire knowledge network structure. Users can see the associations between topics and easily discover the connections between knowledge.

 Canvas Visualization

Basic Memory can generate canvas files to visualize the relationships between concepts and create interactive knowledge maps. Users can ask AI to create knowledge visualizations:

"Create a canvas that shows the connections between my coffee brewing notes."

The AI ​​generates a .canvas file that provides a visual map of knowledge that can be viewed in Obsidian.

Installation and Configuration

Basic Memory is installed using the uv tool:

# Install using uv (recommended) uv tool install basic-memory # Configure mcpServersConfiguration "mcpServers": { "basic-memory": { "command": "uvx", "args": ["basic-memory", "mcp"] }} # Run sync to watch file changes uv tool run basic-memory sync --watch

Integration with Claude Desktop

Basic Memory is ideal for use with the Claude Desktop application:

1.  Edit the Claude configuration file (usually located at `~/Library/Application Support/Claude/claude_desktop_config.json` ):

{ "mcpServers": { "basic-memory": { "command": "uvx", "args": [ "basic-memory", "mcp" ] } }}

2.  Synchronize local knowledge:

# Synchronize local knowledge updates once basic-memory sync # Run real-time synchronization process (recommended) basic-memory sync --watch

3.  In Claude Desktop, AI can now use these tools:

write_note(title, content, folder, tags) - Create or update a note read_note(identifier, page, page_size) - Read a note by title or permalink build_context(url, depth, timeframe) - Navigate the knowledge graph via memory:// URL search_notes(query, page, page_size) - Search the knowledge base recent_activity(type, depth, timeframe) - Find recently updated information canvas(nodes, edges, title, folder) - Generate knowledge visualization

Coming soon: Basic Memory Pro

A standalone GUI application that provides a simplified installation experience and convenient features:

•  One-click installation

• Visual knowledge management interface

• Enhanced visualization tools

• Simplified multi-project management

• Priority Support

With Basic Memory, users can build a knowledge base that grows with each conversation, revolutionizing the way they communicate with AI assistants. No more losing valuable conversation context, all knowledge is stored on the user's computer as a plain text file, providing full control and ownership.

Whether it's personal research, team collaboration or creative projects, Basic Memory can help users build more coherent and valuable AI-assisted workflows.