Review and optimize the MeMesh memory database. Analyzes health score, finds stale/conflicting/redundant memories, shows work patterns, and suggests cleanup actions. Use when asked to "review memories", "check memory health", "clean up knowledge", or "what's in my memory".
Use MeMesh to remember, recall, and manage AI knowledge across sessions. Triggers when the user asks to remember something, recall past decisions, forget outdated info, learn from mistakes, or analyze work patterns. Also triggers proactively when you make important decisions, fix bugs, or learn lessons worth preserving.
Experimental working-model protocol shipped with memesh v4.1. Status — protocol present + instrumented; effectiveness in real usage is being collected (see `memesh patterns`), not yet proven. Suggests a user-as-CTO / Claude-as-orchestrator / background-agents-as-engineers split. Claude routes work by verifiability, dispatches parallel background agents for high-verifiability technical work, and stays foreground only for strategic/understanding work that the user must own. Use as a default for non-trivial software tasks; report back when it helps or doesn't.