token-cost-optimizer

Use before a large Copilot task when model choice, context size, or parallelism could drive up billed usage — estimate cost pressure early and apply Copilot-specific reduction tactics before you run