What problem does it solve?
Many EDA experiment runs are wasted or lead to misleading conclusions because submissions proceed without a disciplined review of route geometry, reroute reports, and timing deltas; this Skill formalizes that preflight review so experiments start from verifiable evidence and explicit acceptance criteria.
Core Features & Use Cases
- Evidence lock: gather case-level TSV/MD summaries, routed DEF occupancy, reroute reports, and timing delta metrics before claiming optimizations.
- Reflection checklist: evaluate backside presence, long-net capture, timing consistency, capacity pressure, and policy-vs-model dominance with a structured pass/fail output.
- Artifact generation and decision: produce a ranked root-cause hypothesis, one-line hypotheses, minimal A/B/C next-step plans with acceptance criteria, and an explicit go/no-go for experiment submission.
- Use case: run this before submitting a new batch of reroute experiments to avoid resubmitting policy-driven artifacts that lack geometric backing or that worsen timing.
Quick Start
Analyze the latest case TSV, routed DEF, reroute report, and timing deltas and produce a preflight_reflection markdown with ranked root causes, A/B/C plans, and a clear go/no-go decision.