AI Agents Inventory and Attributes

The AI Agents Inventory is equivalent to the AI Agents Registry. The two names describe the same governed construct: the authoritative collection in which every AI Agent instance is formally registered. Registration is the act of entering a governed instance into the collection, and it is what makes the collection an authoritative system of record rather than an informal list. Instances are registered either manually, by a practitioner, or through automation that harvests and registers instances from authoritative sources. However registration occurs, a registered instance is a governed record carrying a Semantic Identifier and its governed attribute values.

A Catalog is a distinct construct built over the Inventory/Registry. Where the Registry is the system of record, the AI Agents Catalog is a discovery and exposure surface built on top of it, giving a person a convenient means of finding, learning about, and engaging or working with specific AI Agent instances and the people and technologies related to them. Because every registered instance carries governed attribute values, the Catalog can present and organize instances along those values — for example, the full set of instances alphabetically, grouped by category or other traits held in the data, by owner, by Subject Matter Expert (SME), by date registered, or by date last modified — so a practitioner can locate the right instance and reach its owners, SMEs, and related systems. The Catalog exposes the registered AI Agent instances, and it can expose related data, content, and services at two levels: at the catalog level, across the whole registry, and at the individual instance level, for a single registered AI Agent instance. The Registry governs what is true about each instance; the Catalog makes registered instances and their associated data, content, and services discoverable and consumable.

Beyond locating instances, the AI Agents Catalog is a knowledge management tool: exploring a registered instance teaches a person about it — its identifiers and names, description, owners, Subject Matter Experts, and other governed traits — and lets them traverse its explorable relationships to related Noun Type instances and documents, such as People, Applications, Technologies, and Documents. The value the Catalog can expose is a direct function of the quality of registration: the more complete and accurate the data and information used to register an instance in the Inventory, the more the Catalog has to present and interlink. Because every registered instance is itself a node in the Enterprise Model — the enterprise’s master data and knowledge graph — and every governed relationship is an edge, the more descriptive an instance and the richer its relationships, the more powerful the Enterprise Model becomes, since these inventories are foundational to it. On this governed foundation the AI Agents Catalog can become the basis for higher-order interactive reports, dashboards, and visualizations — for example, heatmaps of strategic disposition across all registered instances, or, for a single instance, interactive node and dependency diagrams that lead outward to other related nodes — making it a critical knowledge user interface into the Enterprise Model.

Every AI Agents Catalog is one member of the enterprise Catalog of Catalogs — the federated discovery surface spanning all Noun Type Catalogs, the discovery-layer counterpart to the Inventory of Inventories that federates their underlying registries. To the extent a Catalog exposes services, those services correlate to the Service Catalog and are governed for delivery through Service Management. Through this federation an AI Agents Catalog does not stand alone: its registered instances and the data, content, and services it exposes become discoverable alongside those of every other Catalog in the Enterprise Model.

The AI Agents Inventory is the AI Agents Registry — the governed record of every AI agent the enterprise owns, operates, or depends on, each formally registered manually or through automated discovery, including by harvesting from the platform-native agent registries maintained by the agent-building platforms an enterprise uses. Those platform-native registries are governed sources that seed the enterprise Registry; they are not the enterprise system of record, because each covers only the agents built within its own platform, keyed to that platform’s own identifiers. The AI Agents Catalog is built over the enterprise Registry to expose every registered AI agent instance for discovery and consumption. At the catalog level, it can expose the full set of available agents, their capabilities, and how to engage them; at the individual instance level, it can expose a single agent’s interface, invocation endpoint, ownership, autonomy level, and operational status, along with the data, content, and services that agent provides. Because AI agents expose invocable services, those services correlate to the Service Catalog and are governed through Service Management, and the AI Agents Catalog participates in the enterprise Catalog of Catalogs alongside every other Noun Type Catalog.