Starlens was born from my own frustration. As a developer, I've starred thousands of repositories on GitHub — AI tools, frameworks, libraries, utility scripts. But when I actually needed to find one, I could only remember "it's a Python library for vector search," completely forgetting its name. GitHub's native Stars list is just a chronologically sorted pile of links. After Astral shut down, this pain became even sharper. I tried manually organizing with note-taking apps, but the maintenance cost was too high and quickly became "organized ruins." The core idea behind Starlens: let your Stars speak for themselves. Instead of chasing perfect manual categorization, we use AI to automatically parse READMEs, generate semantic summaries, and apply intelligent tags. More importantly, through the MCP protocol, Claude Code, Codex, and other AI agents can directly access your collection — your Stars are no longer static bookmarks, but a dynamic knowledge base that AI can query. From a simple search page to a complete ecosystem supporting Web工作台, CLI, and AI Agent integration, every step has been an iteration based on real usage scenarios. If you've ever experienced the "star and forget" dilemma, give Starlens a try. I'd love to hear your feedback and suggestions!

Starlens

AI-powered semantic search for your GitHub Stars

AI-powered semantic search for your GitHub Stars