# Bounds for fixture 7 (slow_burn). # # Ground truth at campaign-level: 1 campaign of 3 observation rows # (one per operational window — recon, exploit, action). A correct # algorithm scores 1.0 across every metric on this fixture. # # Completeness is the load-bearing metric: a clusterer that lets # multi-week silence fragment the campaign tanks completeness (the # one true class is split across the operational windows). The # adversarial recency_decay_clusterer demonstrates this and the # bound below rejects it. # # Campaign-level fixture only — the three DSL actors model the # operator's three operational windows by design. # # Bounds are loose at v1; tighten as the algorithm matures. adjusted_rand_index: min: 1.0 homogeneity: min: 1.0 completeness: min: 1.0 singleton_recall: min: 1.0