SiftPR

AI-powered desktop PR review tool that helps you identify important changes and organize your code review cycles. It will also provide additional suggestions and insights.

Note this thing has been vibe coded to the max, with little or no attention paid to code quality or correctness. Use in full knowledge thereof.

Features

• Desktop App: Native application built with Tauri (macOS, Windows, Linux)
• GitHub Integration: OAuth login to access your repositories
• AI Analysis: Uses AI to organize PR changes into logical categories
• Review Comments: Draft and submit review comments
• Multiple AI Providers: Support for OpenAI, Anthropic, and OpenRouter (bring your own API keys)
• Secure Storage: API keys encrypted with AES-256-GCM

Prerequisites

• Node.js v20+
• pnpm v9+
• Rust (latest stable)
• GitHub OAuth App credentials

Quick Start

1. Install dependencies

   pnpm install

2. Set up GitHub OAuth

   • Create a GitHub OAuth App at https://github.com/settings/developers
   • Set Callback URL to siftpr://oauth/callback
   • Update credentials in src-tauri/src/github.rs

3. Start development

   pnpm tauri:dev

   This will compile the Rust backend and launch the app with hot-reload.

Project Structure

SiftPR/
├── src/                    # React frontend
│   ├── components/         # UI components
│   ├── contexts/           # React contexts
│   ├── lib/                # API client & utilities
│   └── pages/              # Page components
├── src-tauri/              # Tauri Rust backend
│   ├── src/                # Rust source code
│   ├── Cargo.toml          # Rust dependencies
│   └── tauri.conf.json     # Tauri configuration
├── package.json
└── vite.config.ts

Development

# Start Tauri dev mode
pnpm tauri:dev

# Frontend only (UI development)
pnpm dev

# Build production app
pnpm tauri:build

# Type checking
pnpm typecheck

Tech Stack

• Desktop: Tauri v2 (Rust + WebView)
• Frontend: React 18, Vite, TypeScript, Tailwind CSS, shadcn/ui
• Backend: Rust, rusqlite, reqwest, aes-gcm
• Database: SQLite (local, in app data directory)

How It Works

1. Login: Click "Login with GitHub" to authenticate via your browser
2. Configure AI: Go to Settings and add your AI provider API key
3. Review PR: Paste a GitHub PR URL and click "Start Review"
4. AI Analysis: The app fetches PR data and analyzes it with your chosen AI model
5. Review: Navigate through categorized changes and leave comments

Building for Production

pnpm tauri:build

This creates platform-specific installers:

• macOS: .dmg and .app
• Windows: .msi and .exe
• Linux: .deb, .AppImage, .rpm

Output: src-tauri/target/release/bundle/

Security

• GitHub tokens stored encrypted in local SQLite database
• AI API keys encrypted with AES-256-GCM
• No data sent to external servers (except GitHub API and your chosen AI provider)

Disclaimer

THIS SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES, OR OTHER LIABILITY ARISING FROM, OUT OF, OR IN CONNECTION WITH THE SOFTWARE OR ITS USE. YOU USE THIS SOFTWARE ENTIRELY AT YOUR OWN RISK.

License

This project is licensed under the MIT License. See LICENSE.md for details.

TODO

The following feature ideas need implementation:

• Improved agent performance (consider RLM, tools, MCP integration, and scope it to the local repo etc)
• Pre-ingest repo code into vector db and provide as embedding and check what current repo analysis does