Set tenancy, encryption, and purge boundaries

Separate runtime and operator roles for tenant-scoped semantic cache for repeated FAQ paraphrases. Runtime may only perform the narrow actions that cost owner cutting duplicate GPT calls for Acme Corp FAQs needs; operators get audited break-glass with TTL. Encode a negative probe that denies the privilege trick related to cross-tenant near-duplicate question returns Acme's private pricing blurb.

Redact prompts/PII at collection. Secrets enter via a manager or workload identity — never source, fixtures, or exception strings. Incident CACHE-XTENANT-2 should be impossible if these controls hold. Output allowlists and schema checks stay in force on error paths.

cache.purge(tenant="acme")

Rehearse the rollback or kill switch timed against a clock. Record actor, reason, prior revision/secret/flag, and verification query. Invariant reminder: cache key includes tenant_id + policy_version; similarity ≥ 0.92 and risk≠high.

Abuse cases unique to this lab include the privilege path implied by cross-tenant near-duplicate question returns Acme's private pricing blurb. Prove a read-only role cannot mutate. Break-glass tokens expire; leftover tokens fail the drill. Dependency pin/digest story matters when images or models move under you. Document how to rotate the credential that tenant-scoped semantic cache for repeated FAQ paraphrases uses without a full outage window longer than your dual-run plan. Security evidence is part of ship, not an appendix nobody reads.

List network egress destinations and justify each. Ensure debug endpoints are off by default in the shipping config. Verify that error responses do not echo secrets or raw stack frames to clients. For multi-tenant paths, add a cross-tenant probe fixture. Time the rollback drill twice — once with the author, once with a peer. Store the drill transcript beside the threat notes for the incident id. In this chapter the product is tenant-scoped semantic cache for repeated FAQ paraphrases, the human stakeholder is cost owner cutting duplicate GPT calls for Acme Corp FAQs, and the incident id you design against is CACHE-XTENANT-2. Re-state the oracle in your notes — paraphrase hit returns cached answer with cache=HIT; different tenant never hits — and keep the invariant visible: cache key includes tenant_id + policy_version; similarity ≥ 0.92 and risk≠high. Track hit_rate ∈ [0.25,0.55] and cross_tenant_hits == 0 as the scoreboard. Surface under change control: POST /v1/answer. If you only have forty minutes, finish the fixture for cross-tenant near-duplicate question returns Acme's private pricing blurb before polishing UI. Promotion language stays ternary: promote, hold, or roll back based on evidence, not hope.