agent-readiness-hierarchy

Worked Example: Scoring a Real-ish Store

This walks one fictional merchant — Brightwell Home, a mid-size online homeware retailer — through the Agent-Readiness Hierarchy, to show how the scores produce a decision, not just a chart.

Licensed under CC BY 4.0. Maintained by Serafim Tech Limited · serafimtech.io


The scores

Reading bottom-up: Presence 3 · Comprehension 3 · Transactability 1 · Trust 2 · Leverage 3 — written as 3-3-1-2-3.

Layer Score Why
5 · Leverage 3 Promotions, bundles, and loyalty are all exposed in agent-readable form; agent traffic is tracked and attributed separately. Genuinely optimized.
4 · Trust 2 Returns and substitution rules are clear and machine-readable; delegated-authorization handling works but isn’t yet structured to prefer trusted agents. Functional.
3 · Transactability 1 The cart accepts agent line items but fails to return a checkout URL with a correct total when shipping varies by region. An agent can start a purchase but can’t reliably finish one. Basic, not functional.
2 · Comprehension 3 Catalog is clean, every variant resolves, semantic metadata and localization are all in place. Optimized.
1 · Presence 3 Discovery profile is reachable and parseable; a live agent endpoint is published; agentic channels are switched on. Optimized.

How to read it

Bottleneck = Layer 3 (Transactability), at 1. It’s the lowest layer scoring below 2. Everything above it is wasted until this is fixed. Brightwell’s beautiful Leverage layer (5 = 3) earns nothing in practice, because an agent that can’t complete checkout never reaches a promotion.

Overall readiness = Layer 2 (Comprehension). That’s the highest layer scoring ≥2 with no gap beneath it. Presence (3) and Comprehension (3) form an unbroken floor — but the chain breaks at Transactability (1), so readiness stops there. Brightwell is, in effect, a store agents can find and understand but cannot buy from.

This is the trap the hierarchy is built to expose: the headline scores look great (three 3s!) but the store scores as Layer-2-ready, because readiness is gated by your first weak floor, not your best layer.


The fix list this produces

Priority Layer Move Action
1 3 · Transactability 1 → 2 Make checkout return a working URL with a correct total in every region, including variable shipping and tax. This single fix raises overall readiness from Layer 2 to Layer 5.
2 4 · Trust 2 → 3 Structure delegated-authorization signals so trusted agents are preferred, not merely accepted.
3 5 · Leverage hold at 3 Already optimized — protect it; no spend until Layer 3 is fixed.

Note the leverage: because the layers gate each other, fixing one weak floor (Transactability) unlocks the entire stack above it. A 1-point fix at the bottom is worth more than a 1-point gain anywhere else.


The general rule, restated


© 2026 Serafim Tech Limited. Released under CC BY 4.0 — free to use, adapt, and share with attribution.