Beads

An AI-supervised issue tracker designed for coding workflows that provides MCP server integration for task management, dependency tracking, and work discovery. Built by Steve Yegge, it combines SQLite storage with JSONL exports for git-based collaboration, featuring automatic compaction using Claude for issue summarization, daemon-based RPC for performance, and specialized tools for finding ready work, managing dependencies, and tracking discovered issues during development. The system supports multi-machine workflows through git synchronization and includes examples for agent automation, making it useful for AI-assisted development workflows, project management with dependency awareness, and maintaining context across coding sessions.