We extend the Triple Lock Index (Igoshev, 2026) into a three-dimensional phase space defined by stellar spectral class, planetary mass Mp, and relative XUV fluxFXUV. The resulting 5 × 6 × 3 = 90 cells partition the parameter space of potentially habitable worlds. After excluding physically impossible configurations, ∼60 cells re-main; ∼40 % are currently empty in the NASA Exoplanet Archive (February 2026). Empty cells are classified as Type I (physics barrier, TLI < 0.01 %), Type II (ob-servationally inaccessible), or Type III (observable but unoccupied — direct model redictions). A narrow “high-TLI island” (TLI ≥ 1 %) exists for K0–K3 and G-typehosts with Mp = 0.8–5 M⊕ at FXUV ≤ 3× Solar; all other cells fall below 0.1 %. HD 137010 b (K3.5 V, 146 ly) is identified as the sole confirmed exoplanet whosephase-space cell predicts TLIcorr ≥ 0.44 % for all rocky-mass scenarios. Three falsifi-able predictions are issued on 2-, 5-, and 10-yr timescales. All matrix values carry a90 % Bayesian credible interval of TLI/5, TLI × 5, reflecting the N = 1 Earth-only calibration.
Tema Igoshev (Mon,) studied this question.