Stars
Forks
Watchers
Developer links
Shiori
Most web links eventually break - the sobering statistic Shiori, a bookmark manager with archiving by default, is built on. Its answer is archiving by default - where possible, every bookmark you save gets a clean, readable offline copy parsed from the page, ads and navigation stripped, so the article survives even after the original URL dies. Conceived as a simple Pocket clone and written in Go, the entire server is a single binary using roughly 25-30 MB of RAM with SQLite out of the box (Postgres and MySQL supported) - genuinely the lightest archiving bookmark manager you can run. Saving is one click through the Firefox and Chrome extensions, and finding things again is where Shiori quietly outperforms its size: full-text search covers the archived page content, not just titles and tags, so you can find that article by a phrase you remember from paragraph six. Reader mode presents the cleaned text; archive mode shows the preserved page. It's dual-interface by design - a pretty web UI (installable as a PWA on mobile) and a complete CLI for terminal devotees - plus a REST API for scripting. Pocket imports work natively, and Netscape HTML handles browser imports and exports. Multi-user support included. MIT-licensed.
Benefits
- Immune to Link Rot
- Readable offline archives mean the article survives after the original page disappears.
- Search What Pages Said
- Full-text search covers archived content - find articles by phrases, not just titles.
- Absurdly Light
- A single Go binary with SQLite, ~30MB of RAM - runs on the smallest instance.
- Leave Pocket in an Afternoon
- Native Pocket import plus Netscape HTML for browser bookmarks.
Features
- Automatic Offline Archives
- Clean readable copies parsed from every bookmarked page by default.
- Reader and Archive Modes
- Stripped-down article text or the preserved original page.
- Browser Extensions
- One-click saving with tags from Firefox and Chrome.
- Web UI and CLI
- A clean web interface (PWA-installable) plus a full command-line client.
- Full-Text Search
- Query titles, tags, URLs, and archived page content.
- REST API and Multi-User
- Scriptable bookmarking with separate user accounts.