Evaluate with evidence

Track for prompt caching: cache hit rate, $ /day input, p5 time-to-first-token, invalidation lag. Report fractions like 36/40, not vague quality adjectives. Segment by language, plan tier, document length, or other slices that matter for the policy-manual assistant.

Freeze inputs and neighboring versions while evaluating prompt caching. Change one control. Pair results case by case on the policy-manual assistant. Define hard gates (severe errors, privacy, latency) before the bake-off. Use deterministic checks where possible; humans for nuance; model judges only with calibration against gold.

Numeric reminder for prompt caching: without caching: 12.3M input tokens/day; with 90% prefix hits: 1.2M uncached prefix + 0.3M suffix = 1.5M full-price-equivalent tokens before cache-read pricing

Resist adding a twelfth metric before the first three for prompt caching on the policy-manual assistant have owners. This workload improves faster when a small scorecard is trusted than when a warehouse of unused plots exists.

Also pin one numeric memory from this prompt caching chapter: without caching: 12.3M input tokens/day; with 90% prefix hits: 1.2M uncached prefix + 0.3M suffix = 1.5M full-price-equivalent tokens before cache-read pricing That number is not decoration; it is a template for how claims about prompt caching on the policy-manual assistant should look in design docs. Scoped specifically to prompt caching / policy-manual assistant / evaluation.

People confuse prompt caching with neighboring buzzwords when debugging the policy-manual assistant. Before changing prompts, ask whether the broken stage was evidence gathering, the prompt caching judgment itself, validation, or the product action. Fixing the wrong stage creates folklore (“we tried prompt caching and it failed”) that blocks the next team on the policy-manual assistant. Scoped specifically to prompt caching / policy-manual assistant / evaluation.

Write a five-line artifact for this page: goal, inputs, check, owner, stop rule. Invent one fluent failure that the check would catch. Keep details specific to prompt caching rather than generic AI advice.

Write a five-line artifact for this page: goal, inputs, check, owner, stop rule. Invent one fluent failure that the check would catch. Keep details specific to prompt caching rather than generic AI advice.

Write a five-line artifact for this page: goal, inputs, check, owner, stop rule. Invent one fluent failure that the check would catch. Keep details specific to prompt caching rather than generic AI advice.

Write a five-line artifact for this page: goal, inputs, check, owner, stop rule. Invent one fluent failure that the check would catch. Keep details specific to prompt caching rather than generic AI advice.