Fermi-estimation harness for "roughly how many / how much / how long" questions. Forces a six-step discipline (Pin → Decompose → Bound → Combine → Sanity-check → Report) that replaces a single hard number with a product of easier ones, carries an explicit range through every factor, and returns a co
Fermi-estimation harness for "roughly how many / how much / how long" questions. Forces a six-step discipline (Pin → Decompose → Bound → Combine → Sanity-check → Report) that replaces a single hard number with a product of easier ones, carries an explicit range through every factor, and returns a confidence interval instead of a false-precision point pulled from the air. Use when the user asks fo