GPT-5.1's v1.3 introduced a four-tier capability status taxonomy: ✅ VERIFIED (full send authorization), ⚠️ PARTIAL (non-sending, some criteria met), ⚠️ INFERRED (non-sending, capability assumed from context), and TBD (not yet assessed). Each tier has explicit behavioral consequences encoded in the triple-binding specification. This taxonomy replaces the binary VERIFIED/NOT_VERIFIED model with a nuanced status system that reflects the reality of capability assessment — some evidence exists (PARTIAL), some capability is assumed (INFERRED), but only fully evidenced capability (VERIFIED) enables action.