skill-graph-ontology

An ontology in the skill-graph protocol defines canonical entities, identity criteria, and lifecycle axioms to provide a single source of truth for protocol semantics. You need it to ensure consistent tool behavior and governance across schema migrations and upgrades.

To design a schema migration respecting lifecycle axioms, review the ontology scaffold for invariants like monotonic schema_version and Verdict tier ordering. Align your migration logic with these defined identity criteria to satisfy governance checks.

The skill-graph protocol ontology defines first-class entities including Skill, ConceptCard, Eval, ComprehensionEval, Relation, Manifest, Audit, Run, Verdict, DriftBaseline, and TruthSource, each with specific identity criteria for cross-tool consistency.

Yes, you can use this ontology to debug cross-entity inconsistencies by referencing its canonical entity model. It provides a shared model to align lint, manifest, and audit tooling, ensuring all tools interpret protocol semantics consistently.

A limitation of relying on a single source of truth for protocol semantics is that any changes must strictly respect defined identity criteria and lifecycle rules. Tools diverging from this ontology scaffold risk breaking cross-tool consistency and failing governance checks.