AmintaCCCP/GithubStarsManager

What it solves

GithubStarsManager transforms a cluttered list of GitHub starred repositories into a structured, searchable library. It solves the problem of "starring everything and finding nothing" by automating the organization, categorization, and discovery of starred projects using AI.

How it works

The application syncs your starred repositories and Gists via a GitHub token. It then uses AI models (OpenAI, Anthropic, Ollama, or compatible APIs) to automatically generate summaries, tags, and categories for these repositories. For advanced discovery, it integrates with Cloudflare Vectorize to enable semantic search, allowing users to find projects by intent rather than exact keywords. It also provides a dedicated Q&A assistant for individual repositories and an MCP server to allow AI agents like Claude Code or Cursor to access your organized stars.

Who it’s for

Developers and GitHub users who have accumulated a large number of starred repositories and need a better way to to organize, search, and track updates for the projects they follow.

Highlights

• AI-Powered Organization: Bulk AI analysis to auto-generate descriptions, tags, and categories.
• Semantic Search: Natural-language search and a "Find similar repositories" feature powered by vector embeddings.
• Repository Q&A: A commit-pinned assistant for asking focused questions about a specific repository with traceable sources.
• MCP Server Integration: Enables external AI agents to read and search your AI-enriched star library via the Model Context Protocol.
• Release Tracking: A unified timeline for subscribed repositories with smart asset filtering for one-click downloads.
• GitHub Ecosystem Sync: Bidirectional sync with GitHub Lists, fork management with upstream sync, and Gist management with AI summaries.