Convert any code repository into AI-ready context packs in seconds.
🚀 Solve the "AI context too large" problem.
Repo2AI transforms your project into clean, structured, size-limited files that ChatGPT, Claude, Cursor and other AI tools can understand instantly.
Large repositories are hard to use with AI:
- Context too large
- Token limits exceeded
- Files messy and noisy
- AI cannot understand structure
👉 Repo2AI converts your repo into structured, chunked, AI-friendly context packs.
- Feed large projects into ChatGPT / Claude
- Prepare context for Cursor / Copilot
- Understand legacy systems with AI
- Share structured code context with teammates
Instead of dumping raw code, Repo2AI generates:
output/ ├── project-summary.md ├── controllers_01.md ├── services_01.md ├── entities_01.md ├── sql_01.md └── manifest.json
- Structured by code category
- Chunked to fit AI token limits
- Clean and noise-free
- Ready for copy-paste into AI tools
| Feature | Repo2AI | Repomix |
|---|---|---|
| AI-focused output | ✅ | ❌ |
| Code classification | ✅ | ❌ |
| Chunk control | ✅ | |
| Java optimization | ✅ | ❌ |
| GUI support | 🚧 | ❌ |
👉 https://github.com/fichil/Repo2AI/releases/latest
repo2ai scan ./my-projectDownload the executable from Releases.
go install github.com/fichil/Repo2AI@latest
repo2ai scan ./demo
repo2ai scan ./demo --max-size=10mb
repo2ai scan ./demo --format=txt
| Parameter | Description |
|---|---|
| --max-size | Max size per output file |
| --format | Output format (md / txt) |
- Scan repository
- Classify files (Controller, Service, Entity, SQL, etc.)
- Clean & filter irrelevant files
- Split into size-limited chunks
- Generate AI-ready context packs
- Scan Java projects
- Generate AI-ready context packs
- Generate project-summary.md
- Generate manifest.json
- Split context packs by size
- Support Markdown and TXT
- Clean output before generation
- GUI desktop app (Fyne)
- ZIP export
- Improved project summaries
- Spring Boot deep parsing
- Multi-language support (Go / Python)
- Go (CLI engine)
- JavaParser (code analysis)
- Fyne (GUI, planned)
PRs are welcome.
If this project helps you, give it a star.
MIT