Seanium/FeedMe

What it solves

FeedMe is a lightweight, AI-powered RSS reader that eliminates the need for bulky apps or mandatory logins. It transforms traditional RSS feeds into a clean, responsive static page where users can quickly grasp the main points of articles through AI-generated summaries, reducing information overload.

How it works

The project uses a build-time process to fetch content from configured RSS sources and uses a Large Language Model (LLM) to automatically generate summaries for each article. These summaries are then baked into a static website. The system can be automated via GitHub Actions (running hourly by default) or a local Docker container with a cron job to keep the content fresh without requiring a live backend server for the end-user.

Who it’s for

It is designed for users who want a personalized, self-hosted RSS experience that is fast, minimal, and leverages AI to filter and summarize information efficiently.

Highlights

• AI-Driven Summarization: Automatically generates article summaries using a customizable LLM configuration.
• Zero-Cost Hosting: Can be deployed for free on GitHub Pages, Vercel, or Aliyun ESA Pages as a static site.
• Highly Customizable: Users can define their own RSS sources, categories, and AI prompt settings in a YAML configuration file.
• Flexible Deployment: Supports multiple deployment paths including GitHub Actions, Docker, and various static hosting providers.