010 GitHub stars
02Dual-layered extraction using LLM candidates and deterministic validation
03Support for cross-collection Bridge Tags to enable graph traversal
04Domain-specific vocabularies for technical, narrative, and security content
05Automated relevance scoring via the worth_remembering field
06Seamless integration with memory storage workflows and the /learn command