The default baseline table below covers the standard example corpus only. The heavier `example_07_large_scale_routing_no_warm_start` canary remains performance-only and is tracked through targeted diffs plus the conflict/frontier trace artifacts.
Use `scripts/characterize_pair_local_search.py` when you want a small parameter sweep over example_07-style no-warm runs instead of a single canary reading.
The current tracked sweep output lives in `docs/pair_local_characterization.json` and `docs/pair_local_characterization.md`.
Use `scripts/record_iteration_trace.py` when you want a seed-42 vs seed-43 negotiated-congestion attribution run; the current tracked output lives in `docs/iteration_trace.json` and `docs/iteration_trace.md`.
Each scenario entry in `docs/performance_baseline.json` records the full `RouteMetrics` snapshot, including cache, index, congestion, and verification counters.
For the current accepted branch, the most important performance-only canary is `example_07_large_scale_routing_no_warm_start`, which now finishes `10/10/10` after a bounded post-loop pair-local scratch reroute. The relevant counters for that phase are:
The latest tracked characterization sweep confirms there is no smaller stable pair-local smoke case under the `<=1.0s` rule, so the 10-net no-warm-start canary remains the primary regression target for this behavior.
The tracked iteration trace adds one more diagnosis target: `example_07_large_scale_routing_no_warm_start_seed43`. That seed now remains performance-only, and its blowup is concentrated in late iterations rather than the initial basin. In the current trace, seed `43` matches seed `42` through iteration `5`, then spends three extra iterations with `4` completed nets while `net_03` and `net_06` dominate both `nodes_expanded` and `congestion_check_calls`.
