소개
Facilitates deep retrospective analysis by mining past Claude Code sessions and git metadata to surface patterns in coding habits, architectural decisions, and learning trajectories. It employs an iterative discovery process using specific analytical lenses—such as identifying knowledge gaps, mentoring feedback, or project status—to provide developers with a clear view of where they’ve been and how they can improve. By synthesizing cross-project data and semantic search results, it helps users maintain documentation like CLAUDE.md and bridges the gap between past work and future productivity.
