Professional Project Cleaner & AI Context Preparer — Built by Lidprex Labs
Remove node_modules, flatten codebases for ChatGPT/Claude, and ship clean projects to GitHub — in one click.
RepoPrep is a standalone Windows tool that cleans project folders and prepares them for AI workflows. It removes junk files (node_modules, venv, pycache), flattens your entire codebase into a single folder for uploading to ChatGPT or Claude, and reduces project size from 70MB+ down to a few MB.
Built for developers who work with LLMs and large codebases (100,000+ files).
| Version | Size | Requirements |
|---|---|---|
| v2.2.0 (Latest) | 11MB | None — Standalone EXE |
| v1.1.0 (Legacy) | 70MB | Python 3.10+ |
- Before uploading to ChatGPT/Claude — flatten 500 files into one folder instantly
- Before pushing to GitHub — remove build artifacts and secrets
- Before sharing a project — reduce size from 70MB to under 15MB automatically
| Feature | v1.1.0 | v2.2.0 |
|---|---|---|
| File Size | 70MB | 11MB |
| Python Required | Yes | No |
| AI Flatten Mode | No | Yes |
| Languages | English | EN, AR, FR, ZH |
| Startup Speed | 3-4s | 0.5s |
Copies all source files into a single flat folder — removes folder depth so you can upload an entire project to ChatGPT, Claude, or Gemini in one paste.
Automatically removes:
- Dependencies:
node_modules,venv,.gradle,site-packages - Builds:
dist,build,target,bin,out - Caches:
__pycache__,.next,.nuxt,.pytest_cache - Locks:
package-lock.json,yarn.lock,pnpm-lock.yaml - System/Logs:
.log,.tmp,.DS_Store,Thumbs.db
| Mode | What it does | Best for |
|---|---|---|
| Flatten & Prepare for AI | Copies all files into one flat folder | ChatGPT / Claude uploads |
| Smart Clean | Removes junk, keeps folder structure | GitHub pushes |
| Scan Only | Analyzes without touching files | Pre-clean check |
# Run from source
python main.py
# Build .exe (requires pyinstaller)
build.bat- Windows 10 / 11 (64-bit)
- Standalone .exe — no Python or runtime needed
- 4GB RAM recommended for large projects
MIT License — © 2026 Lidprex Labs
"Building products with reputation, not noise."