# Agent interfaces

By Ivan Stegic, 2026-09-25

Low weight and high differentiation. Most sites score nothing here, and in 2026 that is still reasonable — treat this dimension as an opportunity map rather than a deficiency list.

## What it measures

Whether a public API is described by a specification an agent can read. Whether capabilities are discoverable from a well-known path. Whether form fields carry enough labelling that an agent can tell what each one is for. And whether browsing, searching and filtering happen without a challenge blocking the way.

## How this site handles it

Three of these four pass here because the site has none of the thing they measure: there is no API, no advertised capability and no form. That is the standard working as intended rather than a loophole — you cannot fail at markup you do not need. The contact page explains why it offers addresses rather than a form, which is a choice about what an agent can actually act on, not an omission.

## Weight

8 of the hundred points, across 4 tests. The full definitions are in the open
standard at https://readinessindex.io/open-standard