Project Flattener

A web-based tool that helps prepare source code for Large Language Models (LLMs) by combining multiple files into a single text file. Perfect for when you need to share your codebase with AI models like GPT-4, Claude, or Gemini.

Features

• 🗂️ Drag & drop folder support
• 📏 Token limit awareness for various LLMs
• 🧹 Comment stripping for common programming languages
• ⚙️ Configurable file/folder exclusion patterns
• 📊 Real-time token counting
• 💾 Binary file handling options

Supported Models

• GPT-4 (128K tokens)
• Claude 2 (100K tokens)
• Gemini Experimental 1206 (2M tokens)
• Gemini 1.5 Pro (1M tokens)

Usage

1. Open index.html in a modern browser (Chrome/Edge recommended)
2. Select your target LLM model
3. Configure options:
   • Strip comments to reduce token usage
   • Include/exclude binary files
   • Adjust file exclusion patterns
4. Drag a folder or use the "Select Directory" button
5. Copy the processed output to clipboard

Comment Stripping Support

Supports comment removal for:

• JavaScript/TypeScript
• Python
• PHP
• Java
• C/C++
• Ruby
• Go
• Rust
• Shell scripts
• And more...

Default Exclusions

Comes with sensible defaults for excluding:

• Version control directories (.git, .svn)
• Dependencies (node_modules, vendor)
• Build outputs (dist, build)
• IDE files (.idea, .vscode)
• Compiled files (*.exe, *.pyc)
• And many more...

Browser Support

• Chrome/Chromium (recommended)
• Edge
• Other modern browsers with File System Access API support

Local Development

1. Clone the repository
2. Open index.html in your browser
3. No build process or dependencies required

License

MIT License - See LICENSE file for details