Anticipate failure modes

For each severe prompt caching failure on the policy-manual assistant, define stop condition, safe state, owner, and lasting prevention. Rollback only works if prior prompts, indexes, and models remain available. “Send to a human” needs queue capacity and context—not just a button name.

Run one tabletop on the policy-manual assistant for prompt caching: inject a defect, verify detection, contain, recover, and keep the blameless trace.

After the tabletop, store the injected prompt caching defect for the policy-manual assistant as a regression fixture. If the same failure later reaches users silently, your detection story was aspirational. Detection without a fixture tends to rot for prompt caching.

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 / failure-modes.

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 / failure-modes.

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.