Need AI-Assisted OpenAPI-to-EDM Consistency Checking
Need a productized way to check incoming OpenAPI schemas against an enterprise data model so governance teams know where each new API fits or diverges before it ships.
Take Control Of Your Signals — Become a Naftiko Design Partner Today!
Persona Story:
Riley owns API governance on top of a real enterprise data model, and is already running the consistency check by hand against Claude — feeding incoming OpenAPI schemas in, asking whether they fit the EDM, and reading the diff back. It works, but it isn’t a product. The team needs the same check as a governed capability that produces a structured fit / non-fit report with citations into the EDM, plumbed into the design-review workflow rather than living in a chat window.
Problem Context
- Enterprises with mature enterprise data models (EDMs) lack tooling that checks whether new OpenAPI schemas align with the EDM at design time
- Governance practitioners are already running this check informally with Claude in a browser, treating the EDM as the consistency oracle and the LLM as the diff engine
- Manual final checks remain necessary because, in practitioners’ words, “it yet cannot be fully automated”
- A productized version would need to slot into existing platforms (Spectral lint, design reviews, change detectors) rather than replace them
Problem Impact
- Each new API ships with avoidable EDM drift that surfaces as integration pain months later
- Governance leads burn their highest-leverage attention on ad-hoc Claude sessions rather than on policy and outcomes
- The “where doesn’t this fit the EDM?” report stays in a chat window instead of becoming auditable governance evidence
- Other teams cannot reproduce the check on their own APIs because the workflow is one person’s habit, not a capability
Naftiko Today
- Executable YAML capability specs can wrap an LLM-backed consistency check as a first-class, reusable consume + expose, so the workflow becomes a callable capability instead of a manual chat session
- JSON Schema validation gives the consistency-check capability a structured input contract (OpenAPI in, EDM reference in) and a structured output contract (fit report, divergence list)
- Spectral ruleset (15 rules) sits beside the EDM consistency check so governance combines rule-based lint with model-based fit checking in the same pipeline
- External bindings keep the EDM source — published as JSON Schema, RDF, or however the enterprise hosts it — referenceable by the capability without embedding it in code
Naftiko Tomorrow
- OpenAPI-to-Naftiko import (Second Alpha) will give the consistency-check capability a clean on-ramp from any incoming OpenAPI document
- HTML / Markdown format support (Second Alpha) will let the structured fit report render directly into Confluence or an internal portal as a reviewable artifact
- Fabric capability discovery (v1.1) will publish the consistency-check capability across the enterprise so every API team can run it, not just the governance office
- JSON Schema Store publication (GA) will make the consistency-check capability’s input and output contracts a recognized standard other tools can interoperate with