Glossary term
Definition layer
The definition layer is the tier of a software stack that holds a business’s own declarations — which objects exist, how they relate, who may do what to them, and which steps require approval — as versioned files the organization owns, sitting above the runtime layer that executes those declarations and outside any single vendor that happens to ship an engine for them.
Also known as definition layer vs runtime layerruntime layerbusiness definition layer
In practice
Two layers, two different jobs. The definition layer answers what is true about this business: an opportunity has an amount and an account, a sales rep may edit one but never delete it, a discount above thirty percent goes to finance. The runtime layer answers how that gets executed on a Tuesday afternoon under load: storage, query planning, session handling, caching, the permission check on each request, the audit record afterwards. The first is specific to one company and changes when the business changes. The second is generic infrastructure and changes when the engineering does.
Where a layer ends up sitting is not a matter of taste, and the industry has run the experiment several times. Database vendors competed ferociously while SQL, the language itself, stayed public. Container runtimes fought and converged on one open image format. Microsoft opened the Language Server Protocol and every editor adopted it. OpenTelemetry moved observability’s data model to a neutral foundation. The pattern repeats because a layer that many independent parties depend on cannot stay private to one of them without producing exactly the fragmentation that made everyone want the layer in the first place.
AI raises the stakes on which side of that line the business definition falls, because it multiplies the parties reading it. The application reads the definition. So does the agent deciding which tool it may call, the audit system reconciling what happened, the coding assistant proposing next quarter’s change, and whatever analytics tool the finance team bought without asking. If those declarations live inside one vendor’s console, every one of those readers needs that vendor’s cooperation — and the agent, which cannot fill in a support ticket, simply cannot read them at all.
Neutrality is not free, and the trade should be stated rather than assumed. A layer that several parties depend on moves slower than one company’s internal format; compatibility becomes a constraint on every change, and the version everyone agrees on lags the version one vendor could have shipped. Neutrality is also worth nothing on its own: an openly published definition format that only one commercial engine can execute is a portable file and a captive system. The claim is only ever as strong as the weaker of the two halves — a definition you can read and take with you, and a runtime layer you can actually operate yourself.
Where this term is used
Pages and articles that put this term to work.
Product pages
- Platform ObjectOS is the commercial production platform for ObjectStack apps, combining in-app AI, deployment, and team operations with the open, governed ObjectStack runtime.
- For FDEs ObjectStack gives FDEs an open target format and governed runtime for a client-owned open business ontology; ObjectOS is the optional commercial platform for production operations.
- Product tour See how teams use ObjectOS to take an ObjectStack app from a business request to production: typed metadata, a human-reviewed diff, and governed operation.
Articles
- Enterprise-AI-Ontologie: Warum Definition und Runtime offen sein sollten Ontology MCP wurde im Juni 2026 GA: Die Anbieter haben die Agent-Schnittstelle selbst einem offenen Protokoll überlassen. Die Definition folgt. Offen blieb allein die Runtime — dort wohnt Portabilität.
- Offene Unternehmens-Ontologie: Wem sollte die semantische Geschäftsschicht gehören? Zwischen Nov. 2025 und Aug. 2026 haben fünf Plattformen eine semantische Geschäftsschicht ausgeliefert und meist einen MCP-Lesepfad geöffnet — die Definition blieb drinnen. Protokoll offen, Definition geschlossen.
- Warum AI-Agent-Piloten vor der Produktion scheitern: vier fehlende Schichten Ein Agent kann in der Demo Applaus bekommen und später an einer Frage der Rechtsabteilung scheitern. Das Problem ist meist nicht das Modell, sondern fehlende Semantik, Berechtigungen, Genehmigung und Audit.