Metadata-Version: 2.4
Name: mem_learn
Version: 0.1.1
Summary: Memory MCP Server with SQLite backend
Author: XoriieInpottn
License: MIT
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: pydantic>=2.0.0
Requires-Dist: mcp>=1.0.0
Dynamic: license-file
Dynamic: requires-python

# mem_learn - Memory MCP Server

A Model Context Protocol (MCP) server providing long-term memory capabilities with SQLite backend.

## Overview

This MCP server exposes five memory management tools:

- **mem_search** - Search memory by keywords with multi-hop relation expansion
- **mem_read** - Read full content of memory items by ID
- **mem_write** - Create or update memory items
- **mem_link** - Create/strengthen relations between memory items
- **mem_delete** - Delete memory items and their relations

## Installation

```bash
cd mem_learn
pip install -e .
```

## Usage

### As MCP Server

Add to your Claude Desktop config:

```json
{
  "mcpServers": {
    "mem_learn": {
      "command": "python",
      "args": ["-m", "mem_learn"],
      "env": {
        "MEMORY_DB_PATH": "/path/to/memory.db"
      }
    }
  }
}
```

### Environment Variables

- `MEMORY_DB_PATH` - Path to SQLite database file (default: `./memory.db`)

## Memory Types and Usage Guide

See [MEMORY_MANUAL.md](./MEMORY_MANUAL.md) for detailed information on memory types, search strategies, and best practices.
