Aurora RSS Reader

🇨🇳 中文 • 🇺🇸 English

🎉 v0.1.9 Latest Release | 最新版本

Background Summary Queue, Modernized MCP & Deployment Polish | 后台摘要队列、MCP 现代化与部署完善

• 🤖 Background Summary Queue - Unread articles can now enter a saved local summary queue and reuse generated summaries after restart
• 📊 Runtime Status Panel - Added live queue status in settings so you can inspect queued, running, succeeded and failed summary jobs
• 🔌 Modernized MCP Surface - Reworked MCP into split resource tools, added AI/tag/digest workflows, and moved MCP into its own settings section
• 📚 MCP Usability Upgrade - Recommended tools are grouped, legacy aliases are marked for removal, and search now supports cursor pagination
• 🕒 Time Semantics Aligned - MCP list/search defaults now follow the app's current date_range + time_field settings
• 🐳 Docker Backend Port Exposure - Local Docker deployment now exposes the backend directly on 15432, including /api and /mcp

---

Introduction

Aurora RSS Reader is a cross-platform desktop RSS reader integrated with AI translation and summarization features. It supports multiple layout modes, local data storage, and rich customization options.

Current Version: v0.1.9

Key Features

Reading Experience

• Multiple Layout Modes - Support for three-column and single-column layouts
• Group Management - Organize RSS feeds by categories
• Full-text Search - Quick search in titles and content
• Favorites Feature - Bookmark important articles for later reading
• Time Filtering - Filter articles by publication date

Smart Features

• Article Summaries - Automatically generate key points for quick understanding
• Multi-language Translation - Support full-text and title translation for barrier-free reading
• Multi-language Interface - Chinese, English, Japanese, Korean interface support
• Flexible Configuration - Support multiple translation and summarization services

System Features

• Local Storage - SQLite database, offline available
• OPML Import/Export - Easy data migration
• Auto Refresh - Periodically fetch latest articles
• Dark Mode - Support dark and light themes
• RSSHub Support - Extend RSS feed coverage

Interface Preview

Current UI Highlights

Dedicated MCP Settings Background Summary Queue

Scope Summary & Digest RSSHub Preset Selection Markdown Summary Rendering

Media Feed Support

Video Feed Display Image Feed Display

Blog & Article Enhancement

Context Menu & Collections

Feed Context Menu Add to Collection

MCP Status Display

Zotero Integration

Workspace Advanced Search

Quick Start

System Requirements

• Node.js 22
• pnpm 8+

Installation and Running

# Clone the repository
git clone https://github.com/xiongsircool/aurora-rss-reader.git
cd aurora-rss-reader

# Recommended: Node.js backend
cd backend-node
npm install
cd ../rss-desktop
pnpm install
pnpm dev

# Quick start (Node.js backend)
cd ..
chmod +x start.sh
./start.sh

Development Setup (Manual)

Recommended (backend-node):

• cd backend-node && npm install
• cd rss-desktop && pnpm install
• pnpm dev (from rss-desktop)

Troubleshooting

macOS Compatibility

Initial Setup:

1. Download the .zip version if the .dmg fails to mount (common on some systems for unsigned apps).
2. Unzip and drag Aurora RSS Reader.app to your Applications folder.
3. If you see "App is damaged and can't be opened" (or "cannot open"):
   • This is normal for unsigned open-source apps on macOS.
   • Open Terminal and run:

     sudo xattr -rd com.apple.quarantine /Applications/AuroraRSSReader.app

   • Then open the app again.

Update Limitation:

1. The current macOS build is still distributed as an unsigned app.
2. Built-in version checking can work, but automatic in-place installation through ShipIt may fail signature validation.
3. For macOS users, the recommended update path is still to download the latest .dmg or .zip from GitHub Releases manually.

Performance

Q: The app is slow to start? A: Check the backend service logs if possible. First launch may take longer to initialize the database.

• Initialize database
• Launch Electron application

Access URLs

• Desktop App: Electron window opens automatically
• Web Interface: http://localhost:5173
• API Service: http://127.0.0.1:15432

Configuration

Set environment variables for the Node.js backend to configure AI and RSSHub:

# RSSHub
RSSHUB_BASE_URL=https://rsshub.app

# AI Configuration
GLM_BASE_URL=https://open.bigmodel.cn/api/paas/v4/
GLM_MODEL=glm-4-flash
GLM_API_KEY=your_api_key_here

🐳 Docker Deployment

# Quick start with Docker Compose
git clone https://github.com/xiongsircool/aurora-rss-reader.git
cd aurora-rss-reader
docker-compose up -d

# Access Web UI: http://localhost:8080
# Access Backend API: http://localhost:15432/api
# Access MCP Endpoint: http://localhost:15432/mcp

For detailed Docker configuration, see Docker README.

Note: Docker image only supports linux/amd64 architecture.

Build and Release

# Build installation package
chmod +x build-release-app.sh
./build-release-app.sh

Generated files:

• macOS: AuroraRSSReader-Mac-0.1.9.dmg
• Windows: AuroraRSSReader-Windows-0.1.9-x64-Setup.exe
• Linux: AuroraRSSReader-Linux-0.1.9-x64.AppImage

Public Packaging Notes

• Bump rss-desktop/package.json to the target release version before any public packaging or tagging.
• Run the packaging preflight on main first with workflow_dispatch; this validates the build without creating a GitHub Release.
• Only create the public tag after macOS, Windows, and Linux builds all pass.
• Use a fresh tag such as v0.1.9; do not reuse an occupied release tag like v0.1.8.
• Auto-update continues to work only after the public GitHub Release is created with installers and latest*.yml.
• For the current unsigned macOS build, in-app version checking is available, but users should still update via GitHub Releases manually.

📋 Documentation | 文档

• Chinese Version - Complete documentation in Chinese | 中文完整文档

Tech Stack

• Frontend: Vue 3 + Vite + Pinia + TypeScript
• Backend: Fastify + TypeScript + SQLite
• Desktop App: Electron
• Build Tools: electron-builder

Project Structure

aurora-rss-reader/
├── rss-desktop/          # Frontend code
│   ├── src/             # Vue source code
│   └── electron/        # Electron main process
├── backend-node/        # Node.js backend (Fastify)
├── images/              # Image resources
└── start.sh            # Startup script

Support

• Issue Reporting: GitHub Issues
• Feature Suggestions: GitHub Discussions
• Email Contact: 1666526339@qq.com

License

This project uses GNU General Public License v3.0, which is a copyleft open source license requiring derivative works to also be open source.

---

🎯 Future Roadmap

Short-term Plans (v0.2)

• AI Daily Briefing / Scope Summary - Tag digest, feed/group scope summaries, automation and rerun flows are already available ✅
• Podcast Support - Support for audio podcast subscription and playback ✅
• Reading Analytics - Personal reading habits analysis and data visualization
• RSSHub Support - RSSHub URL configuration, mirror switching and RSSHub feed access are already supported ✅
• Browser Extension - Use AI to analyze sites and generate RSSHub subscription rules

Mid-term Plans (v0.3)

• Data Synchronization - Cross-device data sync and cloud backup
• Plugin System - Support for third-party plugin extensions

---

Give it a ⭐ if you find it useful!

Changelog

v0.1.9 (Current Version)

Background Summary Queue + MCP Modernization + Docker Access

• Background summary queue can persist generated summaries locally and reuse them after restart
• Added queue runtime status panel in settings for live inspection of summary jobs
• MCP moved out of AI provider config into a dedicated settings section
• MCP tool surface modernized with grouped recommended tools, deprecation hints and cursor-based search pagination
• MCP time defaults now follow the app's current date range and time field settings
• Docker deployment now exposes the backend directly on 15432 for local /api and /mcp access

v0.1.8 (Previous Version)

Published release baseline before the current packaging cycle

v0.1.7

Digest Intelligence, Tag Workflow Polish & Update Experience

• Digest 2.0 with history, manual regenerate and locale-aligned language
• Latest + Week digest modes
• Analyze-all full pagination fix
• Summary prompting quality upgrade
• Manual update-check entry in settings

Star History

Community

If you want to discuss product ideas, report usage feedback, or join the user community, you can scan the QQ code below.