ZUSE Automat Agent: Empirical Law Discovery in Elementary Cellular Automata

Version v1.9 adds Fase 33 and Fase 34: descriptor-collision auditing and external length-9/10 validation of the T=15 oscillator mechanism. Fase 33 closes the primitive length-8 descriptor universe. Across all binary circular length-8 backgrounds, only two subpatterns_len4 descriptor collisions exist: 00110111/00111011 and 00010011/00011001. Both are already inside the confirmed T=15 set and both preserve the same defect-cycle family. No unseen length-8 background exists, outside rotations of the known representatives, that shares a T=15 subpatterns_len4 descriptor under the same rule. Fase 34 tests whether the T=15 mechanism is confined to primitive length-8 backgrounds. A preflight census finds primitive length-9 and length-10 backgrounds with temporal background period T_bg=3 under rule_73 and rule_109. A targeted validation over 66 such backgrounds, two rules, and 502 non-zero IC words of length 1..8 performs 33,132 rule/background/IC runs. The targeted test finds 90 T=15 detections across 8 external backgrounds: one length-9 background under rule_73, five length-10 backgrounds under rule_73, and two length-10 backgrounds under rule_109. No length-9 background under rule_109 produced a T=15 witness in this test. This shows that the T=15 mechanism is not an artifact of primitive length-8 backgrounds: it generalizes when T_bg=3 is preserved. The compact descriptor from v1.8 remains a length-8 family identifier; extending that descriptor to variable background length remains a separate symbolic problem. ZUSE Automat Agent is a deterministic, policy-driven discovery loop for elementary cellular automata. It builds an empirical atlas of cycle laws, world families, basin fragility, and background-conditioned oscillator dynamics using fixed evaluators and reproducible scripts. The discovery loop itself is deterministic. Language-model assistance was used only after execution for interpretation and documentation, not for world selection, law evaluation, scoring, or acceptance of empirical results. All results are reproducible from deterministic scripts with no stochastic components in the discovery loop. Repository:https://github.com/conchaestradamiguelangel-droid/zuse-automat-agent