1. Artifact claim. This deposit is the Phase II companion preprint (v0.3 PDF): an empirical-and-simulation argument that, in a residual IRGC missile force after high week-one attrition, an open-source attribution-correlation proxy can discriminate active from pre-programmed C2 under sparse provocation—and that dense provocation saturates that proxy (H_ceiling). 2. Problem / scientific context. Parent paper concept 10.5281/zenodo.19210451 (version 10.5281/zenodo.19558494). Phase II tests the residual-force boundary: sparse provocation (April–June 2026) versus dense provocation after July 2026. 3. Contents. Single PDF PHASE2_C2_Observability_REL_FIXED.pdf (v0.3, house typesetting). Not in this deposit: simulation code, CSV, figure scripts (software concept 10.5281/zenodo.19210120). 4. Reproduce. PDF is not executable. Companion repo chokmah-me/milcom-2026-iran-attrition: python verify_milcom_claims.py; simulation/phase2_runner.py; simulation/leg3_null_sim.py. Leg 3 prereg 10.5281/zenodo.21443990. 5. Dual concept. Paper concept 10.5281/zenodo.21865793 ≠ software concept 19210120. Do not merge PDF and zip. 6. Limitations / non-claims. Leg 2 Stream B is small-n (no inferential test on the empirical rate alone). Attribution correlation is a proxy, not the emergent target ratio. High-connectivity separation reflects pure H1/H2 archetypes. Dense July provocation hits H_ceiling — not a standing C2 indicator. Not a MILCOM-acceptance claim. 7. This version (v0.3). Table I H1/H2/H3 D_win levels bound to data_phase2/phase2_discrimination.csv (200-seed). c_res=0.65 H1 D_win is 0.433. House PDF with Hebrew epigraph and date. Replaces the 13 Aug typesetting-incomplete file on 10.5281/zenodo.21925896. 8. Citation. Prefer paper concept DOI 10.5281/zenodo.21865793 for always-latest PDF. Pin this file with this record's version DOI. Software: 10.5281/zenodo.19210120.
command and control · attribution correlation · residual force · launcher attrition · ballistic missile defense · IRGC · Iran · agent-based modeling · OSINT · MILCOM