Why it matters
The whole point of the talk is to stop re-prompting the same context into an agent over and over, which is a direct, hands-on lesson in cutting wasted tokens from real workflows.
The tokenmaxxing angle
The talk is literally about eliminating repeated re-prompting by having a persistent second brain author agent instructions, a concrete pattern for lowering per-task token spend.
From the organizers
Speaker Tommy Puglia bases the demo on his own post, 'Stop Re-Prompting: Your Second Brain Should Write Your Agent's Instructions,' at promptingbi.com.