cyberneticlibrary

Choose response depth and token budget

token-budget-advisorskillsetup L10
Sheshiyer/skill-clusters
What it does

Estimate response depth and offer token budget choices

Best for

Explicit depth control when users want to dial response length before the answer

Inputs
  • · user prompt
  • · target model
Outputs
  • · depth level options
  • · token estimates
Requires
  • · heuristic estimators
Preconditions
  • · prompt provided
Failure modes
  • · heuristic overestimates by >15%
  • · user ignores offer and asks anyway
Trust signals
  • · 85-90% accuracy stated
  • · multiplier ranges tabulated by complexity