MemoVault

A simplified personal memory system for AI assistants, designed for Claude Code integration via MCP.

Features

• MCP Server: First-class integration with Claude Code
• Flexible Backends: Support for OpenAI and Ollama (local) LLMs
• Vector Search: Semantic memory retrieval using Qdrant
• Simple JSON Storage: Lightweight option for basic use cases
• Easy Configuration: Environment-based setup

Quick Start

Installation

pip install memovault

# For local embeddings (optional)
pip install memovault[local]

Basic Usage

from memovault import MemoVault

# Initialize with default settings (reads from .env)
mem = MemoVault()

# Add memories
mem.add("I prefer Python for backend development")
mem.add("My project deadline is March 15th")

# Search for relevant memories
results = mem.search("programming preferences")
for result in results:
    print(result.memory)

# Chat with memory context
response = mem.chat("What language should I use for my backend?")
print(response)

# Save memories to disk
mem.dump("./my_memories")

Claude Code Integration

1. Configure MemoVault in your Claude Code settings:

{
  "mcpServers": {
    "memovault": {
      "command": "memovault-mcp",
      "env": {
        "MEMOVAULT_LLM_BACKEND": "openai",
        "MEMOVAULT_OPENAI_API_KEY": "sk-..."
      }
    }
  }
}

2. Use memory commands in Claude Code:
   • "Remember that I prefer dark mode"
   • "What do you know about my preferences?"

Configuration

Copy .env.example to .env and customize:

# LLM Backend
MEMOVAULT_LLM_BACKEND=openai  # or "ollama"
MEMOVAULT_OPENAI_API_KEY=sk-...
MEMOVAULT_OPENAI_MODEL=gpt-4o-mini

# Embedder Backend
MEMOVAULT_EMBEDDER_BACKEND=openai  # or "ollama", "sentence_transformer"

# Memory Backend
MEMOVAULT_MEMORY_BACKEND=vector  # or "simple"

# Storage
MEMOVAULT_DATA_DIR=./memovault_data

MCP Tools

Tool	Description
add_memory	Store new information
search_memories	Find relevant memories
chat_with_memory	Memory-enhanced chat
get_memory	Retrieve specific memory by ID
delete_memory	Remove a memory
list_memories	Show recent memories
clear_memories	Clear all memories

Architecture

MemoVault/
├── src/memovault/
│   ├── core/           # Main MemoVault class
│   ├── memory/         # Memory backends (simple, vector)
│   ├── llm/            # LLM providers (OpenAI, Ollama)
│   ├── embedder/       # Embedding providers
│   ├── vecdb/          # Vector database (Qdrant)
│   ├── config/         # Configuration management
│   └── api/            # MCP server & REST API

License

MIT