This is an experimental collection of tools to alow LM studio have memory, edit files that has git initialized on them and more. All of the code (including the README) was VibeCoded using Qwen3.6

MemoryDB & File Tools

Give your AI assistant a memory so it remembers what you've discussed, and the ability to edit files directly — all through a simple MCP connection.

All of the code was VibeCoded using Qwen3.6.

---

🧠 memorylite - AI Memory System

Don't you hate it when you have to explain everything from scrach. Let your AI remember conversations, facts, and preferences across sessions using SQLite.

What it does

• Remembers everything you've discussed in a structured way
• Searches by keyword, type, or free-text patterns
• Stores memories with semantic keywords for better recall later
• Fast SQLite backend (replaces the old JSON file approach)

Setup

1. Install dependencies:

   pip install fastmcp

2. Add to your MCP config (mcp.json or equivalent):

{
  "mcpServers": {
    "memorylite": {
      "command": "python",
      "args": ["/full/path/to/memorylite.py"]
    }
  }
}

3. That's it — your AI will now remember things between sessions.

Where memories are stored: ~/.swordmemory/memory.db (SQLite database)

Advanced: Custom Database Location

By default, the SQLite database is created at ~/.swordmemory/memory.db. To use a custom location, add the --path argument to your MCP config:

{
  "mcpServers": {
    "memorylite": {
      "command": "python",
      "args": ["/full/path/to/memorylite.py", "--path", "/custom/path/to/your/database"]
    }
  }
}

Note: If the path ends with / or \, it's treated as a directory and memory.db will be created inside that folder. Otherwise, the path is used as the full file path directly.

---

📝 pythonFileTools - File Editing Tool

Let your AI edit files on disk without copying/pasting full content each time. Tell it to fix a typo or update a config file, and it does it directly.

What it does

• Edit files by specifying which lines/sections to change
• Create new files with specified content
• Read existing files
• Perfect for files that have git initialized on them

Setup

Add to your MCP config:

{
  "mcpServers": {
    "python_file_tools": {
      "command": "python",
      "args": ["/full/path/to/pythonFileTools.py"]
    }
  }
}

---

Quick Start

1. Clone this repo somewhere on your machine
2. Install dependencies: pip install fastmcp
3. Add both tools to your MCP config as shown above
4. Restart your AI frontend (LM Studio, Open WebUI, etc.)

Your AI will now remember things and edit files for you.