A clean, efficient RSS feed aggregator and summarizer powered by AI.

This is a completely rewritten and improved version of the original RSS-GPT project. The code has been refactored for better maintainability, performance, and cleaner repository structure.

Using GitHub Actions to run a Python script periodically: Calling OpenAI API to generate summaries for RSS feeds, and pushing the generated feeds to a separate content branch. Easy to configure, no server needed.

• Clean repository structure: Generated content is stored in a separate content-branch, keeping the main branch clean and focused on code
• Improved workflow: Added cleanup workflow to prevent thousands of unnecessary commits in the main branch
• Refactored codebase: Completely rewritten with better code organization, typing, and error handling
• Enhanced performance: Parallel processing of RSS feeds for faster execution
• Better logging: Comprehensive logging system for easier debugging and monitoring

• Use AI models (supports latest GPT-4o models) to summarize RSS feeds, and attach summaries to the original articles
• Support for custom summary length and target language
• Aggregate multiple RSS feeds into one, remove duplicate articles, subscribe with a single address
• Add filters to your personalized RSS feeds using inclusive/exclusive rules and regex patterns
• Host your RSS feeds on GitHub Pages with clean repository history

1. Fork this repo
2. Add Repository Secrets
   • U_NAME: your GitHub username
   • U_EMAIL: your GitHub email
   • WORK_TOKEN: your GitHub personal access token with repo and workflow scope
   • OPENAI_API_KEY: (Optional) Only needed when using AI summarization feature
3. Enable GitHub Pages in repo settings:
   • Choose deploy from branch
   • Select content-branch (not main)
   • Set the directory to /docs
4. Configure your RSS feeds in config.ini

Edit the config.ini file to add your RSS feeds:

[cfg]
base = "docs/"
language = "zh"  # Target language for summaries
keyword_length = "5"
summary_length = "200"

[source001]
name = "example-feed"
url = "https://example.com/feed.xml"
max_items = "10"
filter_apply = "title"  # Optional: Apply filter to title
filter_type = "exclude"  # Optional: exclude or include
filter_rule = "keyword1|keyword2"  # Optional: regex pattern

You can specify your preferred OpenAI model by setting the CUSTOM_MODEL environment variable in GitHub repository secrets.

• filter_apply: Where to apply the filter (title, description, or both)
• filter_type: Whether to include or exclude matching entries
• filter_rule: Regular expression pattern for matching

Contributions are welcome! Feel free to submit issues and pull requests.

This project is licensed under the MIT License - see the LICENSE file for details.