Memory API MCP Server

🧠 Intelligent Memory Management for Cursor - Automatically save and retrieve code contexts, patterns, and solutions using the Model Context Protocol (MCP).

⚠️ Important Note: This project is not Open Source. All data is encrypted and embedded in a way that only AI can read and interpret the content.

✨ Features

🤖 Automatic Context Saving: Cursor intelligently decides when to save interesting code
🔍 Semantic Search: Find code patterns and solutions using natural language
🏷️ Smart Tagging: Automatic categorization by language, complexity, and reusability
📊 Structured Metadata: Rich context with file paths, complexity levels, and more
🚀 Zero Configuration: Works out of the box with any Memory API instance
🔄 Real-time Sync: Instant saving and retrieval of memories

🚀 Quick Start

Installation

npm install -g memory-api-mcp-server

Setup with Cursor

Install the package globally:
```
npm install -g memory-api-mcp-server
```

Configure Cursor MCP: Add to your ~/.cursor/mcp.json:

{
  "mcpServers": {
    "memory-api": {
      "command": "memory-api-mcp-server"
    }
  }
}

Restart Cursor and start coding!

Auto-Configuration Script

For easier setup, use our configuration script:

npx memory-api-mcp-server configure

🛠️ Available Tools

Core Memory Operations

create_memory - Create new memories manually
search_memories - Search existing memories by content
list_memories - List all memories with pagination
get_memory - Retrieve specific memory by ID
delete_memory - Remove memories

Intelligent Auto-Save

smart_auto_save - Automatically save interesting code contexts
- Detects complex solutions
- Identifies reusable patterns
- Captures bug fixes and optimizations
- Saves configurations and utilities

🧠 Intelligent Auto-Save

The server includes smart auto-save functionality that automatically captures:

✅ Complex Solutions - Code that took significant effort to implement
✅ Reusable Patterns - Functions and classes with high reuse potential
✅ Bug Fixes - Interesting problem resolutions
✅ Optimizations - Performance improvements
✅ Configurations - Useful setup and config code
✅ Utilities - Helper functions and tools

Example Auto-Save

When you write something like this:

const optimizeQuery = (sql) => {
  // Complex optimization logic
  return optimizedSql;
};

Cursor automatically saves it with structured metadata:

{
  "content": "CONTEXTO: optimization\nARQUIVO: utils/db.js\n\nconst optimizeQuery = (sql) => {\n  // Complex optimization logic\n  return optimizedSql;\n};",
  "containerTags": ["optimization", "javascript", "medium_complexity", "high_reusability"]
}

📝 Usage Examples

Manual Memory Creation

// In Cursor chat
"Create a memory about this regex pattern for email validation"

Searching Memories

// In Cursor chat
"Search for memories about database optimization"
"Find all JavaScript utility functions"
"Show me high reusability patterns"

Auto-Save Configuration

Add a .cursorrules file to your project:

# Auto-Memory Rules
Use `smart_auto_save` automatically when detecting:
- Complex algorithms
- Reusable utilities
- Bug fixes
- Performance optimizations
- Configuration patterns

🔧 Configuration

Environment Variables

MEMORY_API_URL - Your Memory API endpoint (default: https://memory-api-mvp.fly.dev)
MEMORY_API_TOKEN - Authentication token (optional)

Memory API Compatibility

This server is compatible with any Memory API that implements the Supermemory AI specification:

POST /v3/memories - Create memories
POST /v3/search - Search memories
GET /v3/memories - List memories
GET /v3/memories/{id} - Get specific memory
DELETE /v3/memories/{id} - Delete memory

🏗️ Architecture

Cursor ←→ MCP Server ←→ Memory API ←→ PostgreSQL + Vector DB

Cursor: Provides the AI interface and code context
MCP Server: Handles protocol translation and intelligent decisions
Memory API: Manages storage and semantic search
Database: Stores memories with vector embeddings for similarity search

🔗 Links

📧 Contact

Email: theefipe@gmail.com

Made with 🧠 by Italo Coimbra