EntroFeed (熵流) is an open-source, AI-powered RSS reader that leverages Entropy (information theory) and Ontology (knowledge structures) with Large Language Models to understand your professional domains and reading preferences. It helps you build an effective information moat through decentralized RSS feeds.
| Feature | Description |
|---|---|
| 🔗 Personalized Subscription | RSS/Feed aggregation with automatic content fetching; supports RSS and RSShub feeds |
| 📊 Multi-dimensional Scoring | Ontology-based article evaluation (recency, authority, relevance, impact) |
| 🔄 Continuous Evolution | Refine evaluation criteria based on your reading habits and event correlations |
| 🌍 Multi-language Support | Full i18n support for Chinese and English |
- Built-in AI chat powered by AntDesign X
- Summarize, translate, and discuss articles with context
- Attach articles to conversations for deep analysis
- Multi-provider support: 8 LLM providers
| Plugin Type | Options |
|---|---|
| LLM Providers | OpenAI, DeepSeek, Zhipu, Moonshot, Baichuan, Tencent, DashScope, Ollama |
| Content Retrieval | requests, Playwright, RSShub |
| Storage | SQLite + ChromaDB vector store |
| Notifications | Slack, Ntfy |
- Clean, modern UI built with React and AntDesign
- Responsive design with fixed-height layouts
- Dark/Light theme support
| Requirement | Version |
|---|---|
| Node.js | 18+ |
| Python | 3.11+ |
| uv | Latest |
| Docker | 20.10+ (optional) |
# Clone the repository
git clone https://github.com/Moon84/EntroFeed_Reader.git
cd EntroFeed_Reader
# Start with default configuration
docker compose up
# Or configure environment variables first
export DEFAULT_LLM_PROVIDER=dashscope
export DASHSCOPE_API_KEY=your_api_key
docker compose upAccess: Open http://localhost:8000 in your browser.
# Install dependencies
make install
# Terminal 1: Start backend (FastAPI on port 8001)
make dev-backend
# Terminal 2: Start frontend (Vite on port 5173)
make dev-frontendAccess: Open http://localhost:5173 in your browser.
| Variable | Description | Default |
|---|---|---|
DEFAULT_LLM_PROVIDER |
Provider: openai, ollama, dashscope |
dashscope |
DASHSCOPE_API_KEY |
DashScope API key | - |
DASHSCOPE_MODEL |
DashScope model | qwen-plus |
OPENAI_API_KEY |
OpenAI API key | - |
OLLAMA_BASE_URL |
Ollama server URL | http://localhost:11434/v1 |
OLLAMA_MODEL |
Ollama model | llama3 |
| Variable | Description |
|---|---|
SLACK_API_TOKEN |
Slack bot token |
NTFY_BASE_URL |
Ntfy server URL |
NTFY_TOPIC |
Ntfy topic name |
| Variable | Description | Default |
|---|---|---|
ENTROFEED_STORAGE_HANDLER |
Storage handler type | sqlite |
DATA_DIR |
Data directory path | ./data |
CONFIG_DIR |
Config directory path | ./configs |
REFRESH_INTERVAL |
Feed refresh interval (minutes) | 5 |
RECENT_HOURS |
Hours for "recent" entries | 36 |
Handlers can also be configured through the web UI at Settings → Handlers.
| Handler | Required Fields |
|---|---|
| OpenAI | api_key |
| Ollama | base_url, model |
| DashScope | api_key, model |
| Slack | token |
| Ntfy | topic |
entrofeed/
├── src/ # Python backend (FastAPI)
│ ├── app.py # Main application & routes
│ ├── backend.py # Business logic layer
│ ├── handlers.py # Handler base classes
│ ├── mcp.py # MCP server implementation
│ ├── scheduler.py # Feed refresh scheduler
│ ├── metrics.py # Prometheus metrics
│ ├── plugins/ # Plugin implementations
│ │ ├── llm/ # LLM provider plugins
│ │ ├── storage/ # Storage handler plugins
│ │ ├── notification/ # Notification plugins
│ │ └── content/ # Content retrieval plugins
│ ├── services/ # Business services
│ │ ├── feed/ # Feed aggregation
│ │ ├── ontology/ # Knowledge base & tagging
│ │ └── recommendation/ # Recommendation engine
│ └── storage/ # Storage implementations
├── frontend/ # React frontend (Vite)
│ └── src/
│ ├── components/ # React components
│ ├── pages/ # Page components
│ ├── hooks/ # Custom React hooks
│ ├── client-api/ # API client functions
│ └── context/ # React context providers
├── configs/ # Configuration files
│ ├── feeds.yml # Default RSS feeds
│ ├── handlers.yml # Handler configurations
│ └── user.md # User profile/interests
├── tests/ # Test suite
│ ├── unit/ # Unit tests
│ └── e2e/ # End-to-end tests
├── docker-compose.yml # Docker composition
├── Dockerfile # Multi-stage Docker build
└── Makefile # Development commands
make install # Install all dependencies
make run # Run backend with uvicorn
make dev-frontend # Start Vite dev server
make dev-backend # Start backend with auto-reload
make build # Build Docker images
make docker-up # Start Docker containers
make docker-down # Stop Docker containers
make unit-test # Run pytest
make clean # Clean data directory# Unit tests
make unit-test
# Or manually
pytest -vvv -cov
# E2E tests (requires running backend)
cd frontend && npx playwright test| Endpoint | Method | Description |
|---|---|---|
/util/list-feeds |
GET | List all configured feeds |
/util/feed-stats |
GET | Get statistics per feed |
/util/list-feed-entries |
GET | List entries (supports filtering) |
/util/list-handlers |
GET | List configured handlers |
| Endpoint | Method | Description |
|---|---|---|
/api/refresh_feed/{id} |
POST | Refresh a specific feed |
/api/delete_feed/{id} |
POST | Delete a feed |
/api/update_feed/ |
POST | Create/update feed |
/api/import_opml/ |
POST | Import OPML file |
/api/export_opml/ |
GET | Export feeds as OPML |
| Endpoint | Method | Description |
|---|---|---|
/api/about |
GET | Get settings & app info |
/api/update_settings/ |
POST | Update settings |
/api/backup/ |
GET | Download database backup |
/api/restore/ |
POST | Restore from backup |
| Endpoint | Method | Description |
|---|---|---|
/api/agent/chat |
POST | Send message to AI assistant |
/api/agent/sessions |
GET/POST | List/create chat sessions |
/api/agent/tools |
GET | List available agent tools |
/api/translate |
POST | Translate text |
/api/llm/status |
GET | Get LLM provider status |
/api/llm/usage |
GET | Get token usage stats |
| Endpoint | Method | Description |
|---|---|---|
/api/search |
GET | Search entries |
/api/recommendations/interest |
GET | Interest-based recommendations |
/api/recommendations/trending |
GET | Trending entries |
/api/recommendations/similar/{id} |
GET | Similar entries |
EntroFeed provides a Model Context Protocol (MCP) server for external AI integration.
# Start MCP server
entrofeed mcp --port 8765
# Or in stdio mode
entrofeed mcp --stdioContributions are welcome! Please read our Contributing Guide for detailed information.
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'feat(scope): add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
Looking for a way to start? Check out issues labeled good first issue!
Please note that this project is released with a Contributor Code of Conduct. By participating in this project you agree to abide by its terms.
This project is licensed under the AGPL-3.0 License - see the LICENSE file for details.
- FastAPI - Modern Python web framework
- React - UI library
- AntDesign - Design system
- ChromaDB - Vector database
- All our contributors and users
- Author: Yinlei Zhang
- Email: wuyuezhang1984@gmail.com
- Issues: GitHub Issues
- Discussions: GitHub Discussions
If you find EntroFeed helpful, please give us a ⭐ on GitHub!
If you find EntroFeed helpful, please give us a ⭐ on GitHub!