What problem does it solve? Centralized, manual data governance becomes a bottleneck as research organizations scale, creating approval queues, shadow datasets, and governance theater where policies exist on paper but are never enforced. This Skill designs a right-sized governance operating model that centralizes policy definition while automating enforcement, so researchers keep shipping without routing around the rules. ## Core Features & Use Cases - Governance spectrum placement: Positions the organization between a lightweight steward council and federated computational (data-mesh) governance, with a recommended blend for research divisions. - Steward taxonomy and council design: Maps business and technical steward roles onto existing researchers, defines council charter, cadence, and decision remit. - Certification over gatekeeping: Replaces pre-publication approval with producer self-certification against centrally defined policies, splitting failures into hard-stop (security, PII, license, retention) versus visibility-only quality gaps. - Global vs domain-local rule split: Minimizes global rules and requires each retained rule to ship with automated enforcement as policy-as-code. - Use Case: A research division lead needs to stand up data governance without hiring a bureaucracy. The Skill produces a governance design doc, steward roster, council charter, certification spec, six-step bootstrap plan, and fitness functions for quarterly review. ## Quick Start Use the research-data-governance skill to design a governance operating model for my research division, including steward roles, a global-versus-local rule split, and a certification spec.
