Linkit: a local-first AI bookmark manager that summarizes and searches your links

Description and use

Linkit turns a static bookmark list into an AI-enhanced, searchable, and maintainable knowledge base, reducing search time and increasing link revisit rates.

Problem

Users accumulate thousands of bookmarks but rarely revisit them because browser bookmarks only store URLs and lack search, summarization, duplicate detection, or health checks.

Outcome

Linkit transforms a static bookmark list into an AI-enhanced, searchable, and maintainable knowledge base, reducing the time spent searching and increasing link revisit rates.

Workflow

• Install Linkit via Homebrew or download binaries
• Add a link through the UI
• Linkit fetches the page and generates an AI summary and key takeaways
• The summary and takeaways are stored in the local database
• Semantic embeddings are created for the content
• Search queries are matched against embeddings for meaning-based results
• Duplicate detection runs in the background and flags near-duplicates
• Link health checks run periodically to flag 404s or changed pages
• Optional Supabase sync can be enabled for multi-device access

Evidence

All bookmarks live in a local SQLite database on your machine. No account required, works fully offline, no telemetry, no third-party data sharing.

Source

A background scan finds 404s and silently changed pages, so dead links don't quietly rot.

Source

If you want multi-device sync, there's an optional Supabase backend with row-level security where every record is bound to your own auth.uid() — including the developers can't read your data.

Source