Scenario: Proof Non-Convergence Under Tight Deadline - Interview Scenario

Scenario

A high-priority protocol property remains inconclusive across multiple engine settings. Release pressure is high, and the team is split between adding assumptions, cutting scope, or accepting bounded-only evidence.

OBSERVED METRIC
Convergence burn-down stalls: inconclusive bucket age grows while helper assertions and runtime budgets increase.

45-MINUTE INTERVIEW FLOW
0-5: define requirement and KPI
5-15: classify first divergence boundary
15-25: identify proving artifact packet
25-35: propose bounded fix with owner
35-45: define validation matrix and rollback

Common traps to avoid

• Declaring closure by timeout reduction without reducing state-space uncertainty.

• Adding broad assumptions to get green status without architecture traceability.

• Skipping vacuity and reachability checks while chasing solver runtime.

Scenario debrief

Score scenario responses on assumption discipline, first-divergence analysis, proof-quality evidence, and release-safe mitigation.

requirements -> model -> proof status -> closure decision

closure confidence trend by risk tier

Debrief prompts

1. Which requirement failed first, and is that path legally reachable?

2. Which modeling assumption most influences this result?

3. What smallest reversible action improves confidence without hiding legal behavior?

Key takeaways

• Tie each proof claim to assumption boundaries and reachability evidence.

• Prefer minimal reversible fixes and preserve legal behavior visibility.

Common pitfalls

• Treating runtime reduction as proof-quality improvement without audits.

• Declaring closure while critical covers remain unreachable.

• Using broad waivers instead of first-divergence root-cause ownership.