Feedlynx

Feedlynx is a minimal, single-user service written in Rust that lets you collect links to read or watch later and exposes them as an Atom/RSS feed. It stores the feed as a file on disk, fetches each URL to extract title/description, and generates YouTube embeds for video links.

Key Features

• Generates and maintains an Atom/RSS feed on disk and serves it over HTTP
• Simple token-based authentication: FEEDLYNX_PRIVATE_TOKEN for writes and FEEDLYNX_FEED_TOKEN for feed path
• HTTP API endpoints: GET / (info page), POST /add (add a link), POST /info (server info), GET /feed/<token> (subscribe to feed)
• Automatic metadata fetching for submitted URLs and YouTube embed generation for video links
• Feed trimming policy keeps up to 50 entries and removes the oldest entries older than 30 days when exceeding the cap
• Distributed as precompiled binaries, buildable with Cargo, and containerisable via the included Dockerfile
• Small footprint and cross-platform: runs on BSD, Linux, macOS, Windows

Use Cases

• Maintain a personal read-later RSS feed to subscribe from any RSS reader
• Capture links from mobile share sheets or browser extensions into a single feed for later consumption
• Integrate with simple automation workflows to collect articles, videos, or other links programmatically

Limitations and Considerations

• Authentication is single-token based; there is no multi-user or per-user account management
• Persistence is a single on-disk Atom file (no database); not intended for heavy concurrent writes or multi-tenant use
• Feed trimming is enforced (30 days / 50 entries) which may remove older items automatically
• The server does not provide built-in HTTPS termination; deployment behind a reverse proxy is recommended for TLS and production setups

Feedlynx is best suited for lightweight, personal read-later workflows where simplicity and portability are priorities. It provides a pragmatic API and simple deployment options for quickly generating a subscribeable feed from collected links.