Tim's Week 4 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 Advisor Strategy Planner
Five questions place your setup on the advisor shape and draft your handoff contract, with the Haiku + Opus cost numbers.
The Human Handoff Audit
Six questions audit your escalation design layer by layer, with the containment math and the plumbing-company proof.
The Production Agent Check
Eight questions check whether your agent is a demo or production-ready, across the four patterns that actually work.
The AXFD Level Check
Five questions, one per stack experience, place your system on the Level 0–4 ladder and name the weakest experience to fix.
The Agent Failure Stack
Six questions, one per failure layer, backed by 73 incidents — with the payment-agent case that ran 34 transactions in 71 minutes.
The Step Reliability Calculator
Slide per-step accuracy and step count to see end-to-end success fall by p^n — and why only shortening moves the math.
The Vendor Call Scorecard
Eight vendor questions scored SHOWN, DESCRIBED, or DODGED — three dodges cap the verdict at DO NOT SIGN YET.
Is Your Eval Setup Measuring the Right Layer?
Break a refund run by switching eval layers off, then check your own setup across outcome, trajectory, efficiency, and safety.
Which Change Strategy Fits Your Team?
Five profile questions across six change frameworks, with the watch-out per framework and the four factors rollouts share.
Can Your Agent Platform Survive a Second Team?
Ten questions, two per guardrail, scored Evans's way — and ‘we think so’ counts as no.
Does Your Agent Drift by Turn 20?
Six questions across the three drift types, plus memory under load and repetition reliability — 90% at turn 1 falls to 54% by turn 20.
Is Your Agent a Demo With the Safety Off?
Eight questions, two per test layer, with the 2am tool-shape story and the pilot math that broke on day 9.
Scorecard
Your ratings, ranked by total. Saved on this device.
Top pick per phase
Full ranking
| Drop | Phase | Uniq | Biz | Total |
|---|