Tim's Week 5 picks
One winner per adoption phase, in phase order — explore, pilot, scale, embed — plus an honorable mention. Tap any row to jump to the card.
The OWASP ASI Exposure Audit
Ten questions, one per ASI risk, each answer revealing the OWASP case and the control that closes the gap — verdict ranks all ten risks most-exposed first.
The Pilot-to-Production Check
Eight questions against the 12-item production-readiness checklist and the four phases: Scope, Evaluate, Integrate, Operate.
The Review Queue Stress Test
Six questions on the six-signal trigger taxonomy, calibration, four-tier routing, and the capacity arithmetic.
The Recurring Agent Monitor
Seven questions on the workflow contract, the canonical run record, run-health alerts, and quality-drift thresholds.
The Eval Predictiveness Check
Eight questions on task set, trace, three measurements in order, repetition, and rerun triggers.
The Approval Fatigue Check
Six questions on decision load, forced thought, the hollow-oversight signal, and skill atrophy.
The MCP Schema Rent Check
Seven questions on the context tax, the accuracy curve, platform ceilings, and the Tool Search fix.
The Multi-Agent Reliability Audit
Six-check audit across the four reliability layers: eval gate, prompt versioning, replayable event log, cost per user.
The Prompt-Change Release Builder
Version the five-part decision set, regression-test against fixed past cases, and ship with a named approver — the record is yours to keep.
The Context-Forgetting Simulator
Break-it simulator for Drew Breunig’s four context failure modes: poisoning, distraction, confusion, clash.
Scorecard
Your ratings, ranked by total. Saved on this device.
Top pick per phase
Full ranking
| Drop | Phase | Uniq | Biz | Total |
|---|