Searches team knowledge cards and session events using keywords and structured attribute filters.
The Session Memory Search skill provides a powerful interface for querying a project's collective intelligence stored within the .claude/memory directory. It allows developers to instantly retrieve technical decisions, 'gotchas,' runbooks, and historical session events using a sophisticated weighted search algorithm. By supporting structured filters for tags, file paths, and importance levels, this skill ensures that critical project context and institutional knowledge are always accessible, reducing duplicate efforts and speeding up developer onboarding within the Claude Code environment.
Key Features
01Dual-layer searching across curated knowledge cards and raw session event logs
02Clear, formatted output providing TL;DR summaries and direct file references
03Recency and importance-based scoring to surface the most critical information first
040 GitHub stars
05Structured filtering by tags, card type, file association, and importance
06Weighted search algorithm prioritizing titles and tags for higher relevance
Use Cases
01Searching through past session events to understand the context of previous implementation changes
02Locating specific architectural 'gotchas' or constraints related to a specific source file
03Retrieving project-specific runbooks and onboarding guides for new team members
