Coder DB - AI Memory Enhancement System

What is Coder DB - AI Memory Enhancement System?

Coder DB is a structured memory system for AI assistants, specifically designed to enhance coding capabilities by integrating multiple database types (Qdrant vector database, SQLite, and a knowledge graph) via the Model Context Protocol (MCP) with Claude Desktop.

How to use Coder DB - AI Memory Enhancement System?

Install dependencies using Poetry, configure environment variables in a .env file (e.g., Qdrant host, port, API key, SQLite database path), and run the FastAPI MCP server. Store code memories using the qdrant-store-memory command and retrieve them with qdrant-find-memories. Query the SQLite algorithm database via SQL or explore the knowledge graph for relationships.

Key features of Coder DB - AI Memory Enhancement System

• Stores code patterns, solutions, and documentation with semantic search.
• Tracks algorithm versions with performance metrics and change logs.
• Represents relationships between coding concepts via a knowledge graph.
• Enables enhanced problem-solving, pattern learning, and project setup workflows.
• Includes security measures (access controls, backup, sanitization).
• Provides usage tracking and performance monitoring.

Use cases of Coder DB - AI Memory Enhancement System

• Query memory for similar solutions when facing a new coding problem.
• Discover and store reusable patterns with metadata and quality metrics.
• Set up new projects with templates, best practices, and boilerplate code.
• Track algorithm improvements and compare alternative implementations.

FAQ from Coder DB - AI Memory Enhancement System

What databases does Coder DB use and where does data live?

It uses Qdrant (vector database), SQLite (relational database), and a knowledge graph. Data is stored locally in Qdrant and SQLite database files; the knowledge graph is also version-controlled.

What are the runtime dependencies and how do I install it?

Requires Python, Poetry, FastAPI, Uvicorn, Qdrant client, SQLite (via aiosqlite), and other dependencies listed in pyproject.toml. Install with poetry install --with dev.

How do I store and retrieve a code memory?

Store with qdrant-store-memory(json.dumps({...})) and retrieve with qdrant-find-memories("query"). Provide metadata like language, complexity, tags, and quality metrics.

What is the MCP server and how does it work?

The MCP server is a FastAPI application that exposes standardized endpoints for AI models (via Claude Desktop) to interact with the memory system. It runs on HTTP and supports health, memory, and algorithm endpoints.

What authentication or security features are available?

Role-based access for sensitive code, optional Qdrant API key (set in .env as QDRANT_API_KEY), and code sanitization to remove sensitive data before storage.