Skip to content

Repository files navigation

AetherFeed app icon

AetherFeed

Native RSS/Atom reader for macOS with AI enrichment: per-article summaries and bilingual tags, category suggestions when adding feeds, and an AI-generated briefing ("What's new") as the home page.

Three interchangeable AI backends:

  • Apple Intelligence — the on-device model (macOS 26+, no setup at all)
  • Ollama — a local or remote Ollama server
  • OpenAI-compatible — OpenAI itself or any compatible server (LM Studio, llama.cpp, vLLM, …)

Features

  • Live-updating article list backed by SQLite (GRDB) with full-text search
  • AI summaries and bilingual topic tags for every article — summaries in German, English, Italian, French, Spanish, or the article's original language
  • Daily briefing home page grouped by your feed categories — regenerated incrementally (only categories with new articles), with an overview that follows your category order
  • Read articles aloud with the system voice (briefing and article view)
  • Ad blocking in the article web view (Adblock-format blocklists)
  • Starred articles, per-feed customization (rename, categories, favicons)
  • Export articles as Markdown (reader-style full-page extraction) or PDF
  • OPML import/export, retention limits, notifications for new articles and the finished briefing
  • Built-in update check against GitHub releases with one-click self-update
  • Localized UI in five languages (English, German, Italian, French, Spanish), switchable at runtime

Build & Run

Requires macOS 15+ and the Xcode Command Line Tools (Swift 6.1+). Xcode itself is not required.

./Scripts/build-app.sh      # builds build/AetherFeed.app (ad-hoc signed)
open build/AetherFeed.app

For development: swift build and run .build/debug/AetherFeed, or ./Scripts/test.sh for the test suite.

Configuration

Settings (⌘,): AI provider (Apple Intelligence / Ollama / OpenAI-compatible) with connection test and model picker, UI and summary language, feed refresh (manually, on launch, or every 5–120 minutes), retention, notifications, update check, and ad-block lists.

License

MIT — © 2026 andeedotnet

About

No description, website, or topics provided.

Resources

Stars

2 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages