Governance Coverage Gaps

A system can correctly apply every rule it knows about and still exercise capability that no rule governs. This paper develops Governance Coverage Gaps (GCG) as a diagnostic framework for that failure. Receipt-bound execution can justify an evaluated action; it cannot by itself establish that every relevant capability was discovered, declared, assigned applicable governance, mediated at grant, observed in use, or represented in the receipt. The framework distinguishes five gap classes: scope-definition, governance-participation, capability-discovery, exception-authority, and evidence gaps. This expansion preserves the source paper's central distinction between governance that exists but fails to participate and governance that was never defined, while making explicit the independent observation needed to discover undeclared capability. It also corrects an important limit: documented non-participation is not automatically intentional or acceptable. An exception removes an obligation only when an authorized, bounded, reviewable rule says so. Governance coverage is formalized as a relation among a bounded capability inventory, applicability rules, required governance layers, evidenced participation, authorized exceptions, and an observation regime. "Full coverage" is therefore never an unqualified claim about all possible capability. It is a closure claim relative to a declared system boundary, discovery method, time interval, threat model, and evidence standard. Integrated with Admissibility Engineering, the Three Hash Rule, the Constitutional Observability Platform, Reasoning by Way of Receipts, Constitutional Computing, Governed Agentic Coding, GMB, and the universal GSM, GCG becomes a method for locating where constitutional machinery is absent rather than merely malfunctioning. This paper supplies a specification and evaluation program, not proof that any implementation has achieved complete governance.