Maintains the accuracy and relevance of AI-recalled context by detecting and remediating stale knowledge.
Sage Knowledge Hygiene is a specialized capability for Claude Code designed to solve the problem of 'knowledge debt' in AI-assisted development. It monitors recalled information for signals of decay—such as version mismatches, outdated temporal markers, and factual contradictions—ensuring that Claude always operates on the most current project context. By providing automated workflows to update, deprecate, or archive information, it streamlines session handoffs and improves the reliability of semantic memory across long-term research and coding projects.
주요 기능
01Proactive knowledge evaluation during recall and session initialization
021 GitHub stars
03Automatic detection of version mismatches and temporal decay signals
04Conflict identification between recalled context and current project facts
05Type-specific decay analysis for todos, preferences, and references
06Remediation tools for updating, deprecating, or archiving stale items
사용 사례
01Cleaning up outdated 'TODO' items from previous research sessions
02Updating project roadmaps after a version release or milestone completion
03Ensuring AI responses do not rely on superseded documentation or API specs
