Threat modelling and adversarial analysis for 0xagentprivacy. Activates when discussing R(t) adversarial bounds, attack surface analysis, red team methodologies, trusted setup attacks, side-channel vulnerabilities, the capability-indexed adversary (moving ceiling, attestation discounting, compositional leakage), or systematic approaches to finding privacy architecture weaknesses. V6 register note (2026-06-10): conjecture and version citations resolve to agentprivacy-docs/research/CONJECTURE_REGISTER_V6.md (head C89); model head: privacy_value_v6_formal_specification.md.
