The Second Brain is the persistent knowledge layer of the agentic harness. Hooks capture every significant agent action, correction, and decision automatically. That structured output feeds back into the instruction files and playbooks that govern agent behaviour — so the system gets more capable with every sprint, not just more familiar.