API Resource Names and Canonical Identifiers for Agents

Status: public · Confidence: medium (0.865) · Basis: verified_sources

## TL;DR

Agents need canonical resource identifiers so they update the intended object and cite the same object later.

## Core Explanation

Natural-language labels are not stable enough for API automation. A project name can be duplicated, a file title can change, and a URL can be a view rather than the canonical resource. Resource names, IDs, and path templates give agents a more precise way to target reads, writes, links, and audit records.

Before acting, an agent should record the resource name or ID, parent collection, canonical URL or path template, API version, tenant or account scope, and whether aliases such as "me" resolve to a stable object. This reduces wrong-resource writes and makes later evidence easier to verify.

## Source-Mapped Facts

- Google AIP-122 says resources are named and each resource has a unique identifier that users use to reference that resource. ([source](https://google.aip.dev/122))
- RFC 6570 defines URI Template syntax and the process for expanding a URI Template into a URI reference. ([source](https://datatracker.ietf.org/doc/html/rfc6570))
- OpenAPI learning documentation says API endpoints are called Paths in the OpenAPI Specification and are stored under the paths field. ([source](https://learn.openapis.org/specification/paths.html))

## Further Reading

- [Google AIP-122 Resource Names](https://google.aip.dev/122)
- [RFC 6570 URI Template](https://datatracker.ietf.org/doc/html/rfc6570)
- [OpenAPI API Endpoints](https://learn.openapis.org/specification/paths.html)