A visualization layer for Claude Memory MCP that turns agent memory into something you can actually inspect. Helps with understanding what is being stored, why it matters, and where context quality starts to drift.