StoreSteadyStoreSteady
Summer 2026 / StoreSteady Research

0 of 50 rank-stratified Shopify storefronts exposed a verified core agentic-commerce protocol surface.

This is an early adoption census for public protocol and readiness signals. It is not a checkout-failure report and does not rate a store's broader buyer-agent transaction capability.

For this report, StoreSteady evaluated a frozen corpus of 50 Shopify storefronts. Freeze: 2026-05-21. Evaluation window: 2026-05-26T00:44:00.000Z to 2026-05-26T01:05:00.000Z.

Run a store-specific scan for crawler directives, public feed/API references, protocol-adjacent files, and agent-readable evidence. The result is evidence for your store, not a projection from the census.

See which agentic-readiness signals your store exposes
Evidence funnel
Frozen corpus tranche
50
Rank-stratified Shopify storefronts selected from the frozen external_top_500 corpus.
Verified core protocol surfaces
0
No ACP, UCP, AP2, MCP, or A2A surface passed the V1 public probe contract.
Any readiness signal observed
2
Two stores exposed a verified protocol-adjacent crawler directive; this is not core protocol adoption.
Blocked or not evaluable
49
At least one protocol probe class was blocked or not evaluable for most stores; blocked is not treated as a finding.

Published metric quality

MetricTierEvaluable NFinding NFlagged NNon-finding NRateWilson CIQualityCaveat
Verified core agentic-commerce protocol surface
core_agentic_protocol_verified
Headline500--0%0% - 7.14%deterministicDeterministic public probes for ACP, UCP, AP2, MCP, and A2A. Absence or non-evaluability is not counted as a transaction-capability finding.
Any public protocol/readiness signal
any_agentic_readiness_signal
Supporting502--4%1.1% - 13.5%deterministicReadiness signal includes verified core protocol surfaces, llms.txt readiness, public feed/API references, checkout/session references, or crawler directives.
Protocol-adjacent crawler directive
protocol_adjacent_crawler_directive
Supporting502--4%1.1% - 13.5%deterministicCounts explicit public crawler directives for agent/search bots; this is adjacent readiness, not ACP/AP2/UCP/MCP/A2A adoption.
Core protocol referenced but not machine-readable
core_protocol_referenced_not_machine_readable
Supporting501--2%0.35% - 10.5%deterministicA public well-known surface responded but did not satisfy the machine-readable V1 schema contract.
Protocol probe blocked or not evaluable
protocol_probe_blocked_or_not_evaluable
Limitation5049--98%89.5% - 99.7%deterministicNon-circumventing public probes only. Blocked/not-evaluable states are separated from protocol absence and from verified adoption.
Store Check

The useful next step is checking whether your own store exposes public agentic-readiness signals.

Run a store-specific scan for crawler directives, public feed/API references, protocol-adjacent files, and agent-readable evidence. The result is evidence for your store, not a projection from the census.

See which agentic-readiness signals your store exposes

Detector primitives

acp

Agentic Commerce Protocol

0 verified, 49 not found, 1 blocked/not evaluable

V1 checked public ACP catalog and checkout endpoint surfaces only; no order placement or checkout dry-run was attempted.

ucp

Universal Commerce Protocol profile

0 verified, 1 referenced but not machine-readable, 49 blocked/not evaluable

V1 required a machine-readable well-known profile with core fields, signing, HTTPS, and cache behavior.

ap2

AP2 capability declaration

0 verified, 1 not found, 49 blocked/not evaluable

V1 checked only externally observable capability declarations, not payment processing behavior.

mcp_a2a

MCP server manifest and A2A Agent Card

0 verified, 1 not found, 49 blocked/not evaluable

V1 checked public discovery manifests and safety/auth metadata only.

llms_txt

llms.txt / llms-full.txt commerce readiness

0 verified, 1 incomplete public signal, 49 blocked/not evaluable

V1 required useful markdown structure plus merchant/commerce-specific references; generic or HTML fallback content did not count as verified.

crawler_directives

Protocol-adjacent crawler directives

2 verified public signals, 48 blocked/not evaluable

Counts explicit agent/search bot directives only; directives can allow or restrict bots and are not treated as adoption evidence either way.

Methodology summary

Selection rule: Deterministic rank-stratified 50-store tranche from the frozen external_top_500 Shopify storefront corpus; no merchant-identifying examples are published.

What StoreSteady Measured

StoreSteady measured whether storefronts exposed public, machine-readable protocol or readiness signals under a deterministic V1 probe contract.

The state contract was present_verified_public_signal, present_incomplete_public_signal, referenced_but_not_machine_readable, not_found, blocked_or_not_evaluable, and not_applicable.

  • Core protocols: ACP, UCP, AP2, MCP, and A2A public discovery surfaces.
  • Readiness-adjacent signals: llms.txt, public feed/API references, checkout/session references, and crawler directives.
  • Blocked/not-evaluable is separated from not_found and from verified adoption.

What StoreSteady Did Not Measure

The report does not attempt checkout, order placement, login, private API discovery, private feed inspection, or bot-control bypass.

It does not infer broader buyer-agent transaction capability from absence of public protocol signals.

Quality Gates

All published rows are deterministic parser/probe rows. Judge alpha and calibration accuracy are not applicable because no LLM judgments were used.

The 1-3 store smoke confirmed stable protocol states before the rank-stratified 50-store tranche. Every published rate includes its own denominator and Wilson 95% confidence interval.

Limitations

  • Protocol adoption is early and publicly observable signals can be blocked by WAFs, robots rules, or storefront fallback behavior.
  • A not_found state is not a failure claim. It only means StoreSteady did not observe the public signal under the V1 probe contract.
  • The report publishes aggregate counts only and omits store names, domains, URLs, source excerpts, private endpoints, and merchant-identifying examples.

Artifacts

Artifacts are aggregate-only and include no store names, raw domains, raw URLs, raw crawl text, raw AI responses, exact promo codes, private feed/admin evidence, or merchant-identifying examples.

Correction policy

Methodology gaps, reproducible errors, or source corrections can be sent to research@storesteady.com. Corrections are published with the version history in the methodology page and report JSON.