Platform: EnergyScope MCP v2.9.3
Pipeline: data → derived_signals (v2, scored/themed/transition-aware) → compare_runs → HTML → publish
Tools used: latest(daily), derived_signals, compare_runs, upcoming_events, IEA.OMR_HEADLINE
Data sources: ICE, SHIP, MACRO, GAS, CFTC, BH, IEA
Contradictions firing: diesel_clean_long (1.000), forced_storage_build (0.965), supply_shock_signature (0.922), regime_curve_destruction (0.920), freight_locked_arb (0.725), fundamental_rally (0.725)
Snapshot keys: wti, brent, ice_m1, ice_m12, m1_m12, roll_yield_pct, bwet_z, brent_wti_zscore, crack_321_z, gasoline_seasonal_pctl, production_seasonal_pctl, wti_mm_pctl, brent_mm_pctl, vix_corr, cushing_z, cushing_7d_pct, wti_price_pctl, nymex_wti, ttf_eur_mwh, iea_global_supply_mbd, iea_hormuz_loss_mbd, iea_demand_revision_kbd, bh_us_rigs, regime, primary_dislocation, dominant_lens, contradictions_active_count, signals_active_count
Signal layer: 45 signals evaluated, 19 active across 6 themes. Dominant lens: diesel_clean_long (1.0). Virtual metric: OPEC-IEA spread +2.53 mb/d.
Narrative approach: Regime-first hierarchy. Contradictions are server-canonical. Replication: any MCP client calling the same tools produces identical analysis.