An RSS reader that uses AI to do the actual reading

What it does

MrRSS is a cross-platform RSS reader that auto-translates and summarizes articles, then pushes them to tools like Obsidian, Notion, FreshRSS, or RSSHub. It ingests standard feeds, XPath scrapes, newsletters, and custom scripts, wrapping the logic in a Go backend with a Vue.js face. If you prefer a remote headless setup, it also runs as an API-only server with a Swagger spec.

The interesting bit

The app treats subscriptions as a data pipeline rather than a reading list: built-in filters and a scripting system automate what gets ingested, summarized, and forwarded elsewhere. It is also built on Wails v3, which is still in alpha — a choice that is either admirably current or slightly brave.

Key highlights

• Auto-translates titles and content, then generates summaries via unspecified “advanced AI technology.”
• Native plugins for Obsidian, Notion, FreshRSS, and RSSHub.
• Handles standard URLs, XPath, scripts, and newsletters alongside typical RSS/Atom feeds.
• Desktop builds for Windows, macOS, and Linux, plus a portable mode that keeps all data in a single data/ folder.
• Headless server mode with pre-built Docker images and a Swagger API for remote deployments.

Caveats

• Built on Wails v3 alpha, so expect framework turbulence.
• The README advertises AI features but never names the provider, model, or whether local inference is an option.
• Scripting and automation are billed as “highly customizable,” yet the README omits what language the scripts use or how they are sandboxed.

Verdict

Worth a look if you want a feed aggregator that summarizes and syncs to your note-taking stack. Look elsewhere if you need a stable, mature reader without alpha dependencies or opaque AI hooks.