Checklists

How to use this resource

Use these checklists before mocks, after mocks, and before scheduling real loops.

Before a mock

• Name the round type, prompt source, timebox, and allowed tools.
• Choose one primary signal and one visible behavior to practice.
• Select a reviewer or recording artifact.
• Prepare a place to capture decisions, misses, and timestamps.
• Define the stop condition; do not turn one mock into an endless repair session.

After a mock

• Write the prompt and major constraints from memory.
• Score visible behavior, not private intent or final correctness alone.
• Preserve the artifact: code, diagram, notes, recording, or transcript.
• Identify the earliest observable cause of the largest miss.
• Write one correction rule and schedule a different retest prompt.

Weekly preparation review

• Every weak signal has a specific repair route.
• Repeated mistakes are tracked across attempts rather than rediscovered.
• Practice includes both isolated drills and complete round simulations.
• At least one delayed retest checks whether the correction transferred.
• The next week removes low-value volume before adding new material.

Before a real loop

• No target-role core lane remains below the agreed readiness threshold.
• Two current project stories survive detailed follow-up probes.
• Coding workflow exposes an invariant and derives tests from risk.
• System design includes failure, observability, rollout, and evolution without prompting.
• Behavioral answers contain specific personal action, evidence, and reflection.
• Logistics, accommodations, tools, time zones, and contact paths are confirmed.
• The final 24 hours protect sleep and execution rather than adding topics.

Gate rule

A checked box needs evidence. Link it to a scorecard, artifact, or recent attempt. If a critical item fails, choose the smallest repair and retest; do not check it because the interview date is close.

For round-specific day-of lists, use Appendix L — Day-of Checklists.