Decision-OS V7: Aspire Intelligence for AGI

Decision-OS V7 proposes a minimal, operational definition of Artificial General Intelligence (AGI) that does not rely on benchmarks or capability-based evaluations.Instead, AGI is defined structurally as the conjunction of three requirements: Aspire (long-horizon direction), Guard (non-negotiable boundary constraints), and Intelligence as stable self-recursion grounded in canonical memory (PIC). Evolution is activated only when the threshold condition ΔR_total > θ_evo is satisfied, and long-run stability is enforced by the idempotent self-update condition F(F(x)) = F(x).These conditions are summarized by the door formula: AGI ≜ A × G × I. In the Decision-OS lineage, this also fixes the structural continuity V4 × V5 × V6 = V7, where V4 contributes Aspire, V5 contributes Guard, and V6 contributes PIC.This work does not propose a benchmark or performance checklist. It instead defines the minimal structural conditions required for sustained self-evolution to remain meaningful, guardable, and non-divergent over long horizons.This item is the primary V7 paper. For supplementary clarification on why Aspire and PIC are necessary, see the related V7 Addendum.