This is a streamlined "Repoprompt" tool. You can use this with o1-pro or o1 to have it add new features to an existing codebase, and then use o1-xml-parser to update the files. I've worked on this to generate input prompts to feed into o1-pro (or o1), and then use (my version)[https://www.github.com/adrianscott/o1-xml-parser] of McKay Wrigley's o1-xml-parser to parse the XML output and update local files. By default, This repo has file-writing turned off for security, since we use the o1-xml-parser to update the files. I've sought to trim this down to only what's needed for o1-pro and o1-xml-parser to work.
This is forked from Justin Lietz's original project: https://github.com/justinlietz93/RepoPrompt-Windows-Linux
From Justin: Feel free to do whatever you want with this code, but please share it and feel free to give me credit. :) This project is still in progress. Most of the functionality works, but you may run into bugs and I have more features planned.
- Clone the repository:
git clone https://github.com/adrianscott/freepoprompt.git
cd freepoprompt- Install system dependencies:
# Ubuntu/Debian
sudo apt-get install python3-tk
# Fedora
sudo dnf install python3-tkinter
# Arch Linux
sudo pacman -S tk
# macOS (if not already installed)
brew install python-tk- Install Python dependencies:
pip install -r requirements.txt- Start the application:
streamlit run main.pyto run on another port, use e.g. streamlit run main.py --browser.serverPort 8502
- Using the interface:
- Select repository using the visual file browser
- Configure ignore patterns if needed
- Add rules/instruction files
- Click "Generate Prompt" to analyze
- Use the Settings tab to:
- Select token analysis model
- Set repository path via browser or manual input
- Use the Files tab to:
- Upload configuration files (.yaml, .yml)
- Upload rule files (.txt, .md, .cursorrules)
- View and manage loaded files
- Use the Patterns tab to:
- Configure ignored directories
- Set ignored file patterns
- Download current configuration
Repository_Crawler/
├── frontend/ # UI components
├── backend/ # Core functionality
├── config/ # Configuration files
├── logs/ # Application logs
├── prompts/ # Generated output
└── tests/ # Test suite
- Repository traversal system needs improvement
- XML generation formatting issues
- UI component state persistence
- Linux / Mac compatibility verification needed
- Fix repository traversal system
- Improve XML generation
- Enhance error handling
- Add comprehensive testing
- Test cross platform compatibility
If you find FreepoPrompt helpful, consider supporting its development:
bc1qvzqjyrmu0tz75g67mfprfsc5gp304nzcvtz83n
(or DM on X for Zelle)
Your support helps maintain and improve FreepoPrompt for everyone!
If you use FreepoPrompt in your research or project, please cite it as:
@software{freepoprompt2025,
author = {Adrian Scott},
title = {FreepoPrompt: Cross-Platform Prompt Engineering Tool},
year = {2025},
publisher = {GitHub},
url = {https://github.com/AdrianScott/freepoprompt}
}The following docs may not be up to date:
A powerful Python-based tool that quickly produces context prompts for LLMs by analyzing local repositories and generating structured documentation with token cost estimation. Built with Streamlit, it provides an intuitive web interface for exploring codebases and understanding their token usage in the context of AI language models.
- Interactive repository exploration
- Token analysis with GPT model support
- Cross-platform compatibility
- Configurable ignore patterns
- XML-formatted codebase overview
- Tabbed sidebar interface for better organization
- Settings: Model and repository configuration
- Files: Upload and manage system files
- Rules: Add rules and instruction files
- Patterns: Configure ignore patterns
- Limited Visual file browser for repository selection
- File tree visualization
- Token analysis visualization
- Extensive ignore patterns for files and directories
- Multiple GPT model support (TODO)
- GPT-4 (default)
- GPT-4-32k
- GPT-3.5-turbo
- GPT-3.5-turbo-16k
- Cross-platform path handling
- Configurable logging system
For support, please open an issue in the GitHub repository (or contact on X).