Fictitious sample · made-up data
HOSHIRO ARC
Agent Readiness and Compatibility Report
An evidence-based assessment of how AI systems and autonomous agents discover, understand, trust and act on this website.
ASSESSED PROPERTY
willow-thread.example
FICTITIOUS ASSESSMENT EXAMPLE
MANUAL VERIFICATION PENDING
Prepared for the named recipient. Contains assessment findings and indicative remediation scope; not a commercial agreement. © Hoshiro Labs.
Fictitious sample · made-up data
01
Assessment status
AUTOMATED ASSESSMENT
Complete
42 automated ARC checks were executed against the sampled pages on 12 October 2026. Every result in this report comes from that executed evidence.
MANUAL VERIFICATION
Pending
6 advanced checks require operator verification and remain not verified. No result has been recorded for them; none is scored or implied as a pass.
Automated assessment complete. Certain advanced checks require operator verification and remain pending. Where this report is silent on those behaviours, it is because they have not yet been observed — not because they passed.
CONTENTS
ARC score & category overview
Result & priority distribution
Manual verification status
Prioritised remediation roadmap
Commercial & scope summary
What ARC tested — coverage
Appendix A — test manifest
Appendix B — page evidence
Methodology, versions & limitations
HOW TO READ RESULT STATES
PASS
The check ran and the expected signal was present.
NEEDS ATTENTION
The check ran and returned a partial or informational result.
FAIL
The check ran and the expected signal was absent or blocked.
NOT VERIFIED
No reliable way to verify existed; never counted as a pass or a fail.
NOT APPLICABLE
The check does not apply to this site; excluded from scoring.
PENDING
Operator verification required; not yet performed.
Fictitious sample · made-up data
02
Executive summary
OVERALL ARC SCORE
54/100
Maturity band C — Developing. Scored across 7 ARC categories by scoring engine v2.0.0.
willow-thread.example has significant agent readiness gaps that likely block AI-mediated discovery, understanding, or action; 4 remediation opportunities identified.
AUTOMATED
Complete · 42 checks
MANUAL VERIFICATION
Pending · 6 checks
STRONGEST AREAS
Business Understanding & Answerability · Conversion & Agentic Commerce
HIGHEST-PRIORITY BLOCKERS
CRITICAL
Site is not globally blocked
Global crawling is blocked site_not_globally_blocked
HIGH
Entity schema exists
Organization/LocalBusiness entity schema missing entity_schema_exists
HIGH
Valid JSON-LD exists
Types: none valid_json_ld_exists
REMEDIATION OPPORTUNITIES
4
2 items carry an indicative price; 2 require a scope review.
Indicative timeline covers scoped items only; remaining items subject to scope confirmation
INDICATIVE REMEDIATION VALUE
USD 4,200.00
Covers 2 priceable items. 2 custom-scope items are excluded from this figure.
ASSESSMENT COVERAGE
ARC's methodology defines 50 checks. This report ran 42 automated checks; 34 produced a scored result, 5 were not applicable and 3 could not be verified. 6 operator-verified checks remain pending. Not applicable and not verified checks carry no points and are not failures.
Categories without a numeric score and why:
AI Visibility & Recommendation (PENDING MANUAL) — it depends on operator verification that is still pending; Identity, Authentication & Agent Access (NOT APPLICABLE) — every check that ran was not applicable to this site.
HOW TO READ THE SCORE
The 54/100 figure is the canonical ARC score across all 7 categories.
2 categories carry no scored automated evidence
—
AI Visibility & Recommendation (pending manual),
Identity, Authentication & Agent Access (not applicable).
Across the 5 categories with evaluated evidence the arithmetic mean is 76 — shown for transparency, it is not an ARC score.
Fictitious sample · made-up data
03
ARC score & category overview
Each category is scored only from checks that actually ran and could be scored. Where nothing was scored, a state is shown instead of a number — a category with no evaluated tests is not a zero-score failure.
CATEGORY
SCORE
EVIDENCE
STATE
Discovery & Crawl Access
9 / 11 passed
SCORED
Machine Readability & Content Structure
5 / 10 passed
SCORED
Business Understanding & Answerability
5 / 6 passed
SCORED
AI Visibility & Recommendation
1 manual only
PENDING MANUAL
Agent Interfaces & Actionability
2 / 4 passed
SCORED
Identity, Authentication & Agent Access
2 not applicable
NOT APPLICABLE
Conversion & Agentic Commerce
2 / 3 passed
SCORED
ARC PILLARS
ARC reports REACH (can AI systems find and understand you) and CAPABILITY (can an agent act with you) as separate axes; they are never blended beyond the overall score.
SCORE COMPOSITION
All categories are weighted equally.
2 categories contributed no scored evidence and are currently counted as 0 by scoring engine v2.0.0, which lowers the overall figure. Mean of the 5 evaluated categories: 76. Both numbers are shown so the score cannot be misread.
Fictitious sample · made-up data
04
Result & priority distribution
AUTOMATED TEST RESULTS · 42 EXECUTED
Not verified and not-applicable results are excluded from scoring entirely — never a silent pass, never a fail. A further 6 operator-verified checks are reported separately and are not counted in the 42.
MATERIAL FINDINGS BY SEVERITY · 11 TOTAL
CRITICAL
1
HIGH
2
MEDIUM
1
LOW
7
REMEDIATION PRIORITY · 4 GROUPED OPPORTUNITIES
CRITICAL
AI Crawler Access & Discoverability
R-01
HIGH
Structured Data & Entity Implementation
R-02
LOW
AI Agent Action & Conversion Readiness
R-03
LOW
Agent-Friendly Documentation & Discovery
R-04
COVERAGE
9 of 12 sampled pages evaluated
Partial coverage — one or more pages in the sample could not be reliably assessed; treat findings as directional for those pages. 3 pages were unreachable during the audit — see appendix B.
DETECTED PLATFORM
Unknown / Manual Assessment Required
No CMS or platform fingerprint was confidently identified. This is a factor in why 2 remediation items are classified Custom Scope Review Required rather than priced automatically.
Fictitious sample · made-up data
05
Findings & evidence
Every finding below is a check that actually executed. Each carries its stable ARC test ID so it can be re-tested and verified after remediation.
F-01 · DISCOVERY & CRAWL ACCESS
CRITICAL
FAIL
Site is not globally blocked
WHAT ARC FOUND
Global crawling is blocked
WHAT SHOULD HAPPEN
Publish or correct robots.txt and remove unintended global or AI-crawler blocks.
CATEGORY
Discovery & Crawl Access
ARC TEST ID
site_not_globally_blocked
COVERED BY
Remediation item R-01 — see the prioritised roadmap.
F-02 · BUSINESS UNDERSTANDING & ANSWERABILITY
HIGH
FAIL
Entity schema exists
WHAT ARC FOUND
Organization/LocalBusiness entity schema missing
WHAT SHOULD HAPPEN
Publish Organization/LocalBusiness schema matching authoritative business details.
CATEGORY
Business Understanding & Answerability
ARC TEST ID
entity_schema_exists
COVERED BY
Remediation item R-02 — see the prioritised roadmap.
F-03 · MACHINE READABILITY & CONTENT STRUCTURE
HIGH
FAIL
Valid JSON-LD exists
WHAT ARC FOUND
Types: none
WHAT SHOULD HAPPEN
Add and validate schema.org JSON-LD covering business entity identity and core offers.
CATEGORY
Machine Readability & Content Structure
ARC TEST ID
valid_json_ld_exists
COVERED BY
Remediation item R-02 — see the prioritised roadmap.
F-04 · MACHINE READABILITY & CONTENT STRUCTURE
MEDIUM
FAIL
Service/Product/Offer/FAQ schema exists
WHAT ARC FOUND
Service/Product/Offer/FAQ schema missing
CATEGORY
Machine Readability & Content Structure
ARC TEST ID
service_product_offer_faq_schema
COVERED BY
Remediation item R-02 — see the prioritised roadmap.
F-05 · AGENT INTERFACES & ACTIONABILITY
LOW
FAIL
API Catalog is discoverable
WHAT ARC FOUND
No API Catalog was found.
CATEGORY
Agent Interfaces & Actionability
ARC TEST ID
api_catalog_available
COVERED BY
Remediation item R-04 — see the prioritised roadmap.
F-06 · AGENT INTERFACES & ACTIONABILITY
LOW
FAIL
OpenAPI documentation is valid
WHAT ARC FOUND
No parseable OpenAPI/Swagger document was found.
CATEGORY
Agent Interfaces & Actionability
ARC TEST ID
openapi_documentation_valid
COVERED BY
Remediation item R-04 — see the prioritised roadmap.
F-07 · CONVERSION & AGENTIC COMMERCE
LOW
NEEDS ATTENTION
Pricing or offer context is visible
WHAT ARC FOUND
No public pricing/offer language detected (may be an intentional sales-led model, not a defect)
CATEGORY
Conversion & Agentic Commerce
ARC TEST ID
pricing_offer_context_visible
COVERED BY
Remediation item R-03 — see the prioritised roadmap.
F-08 · DISCOVERY & CRAWL ACCESS
LOW
NEEDS ATTENTION
AI-specific meta directives detected
WHAT ARC FOUND
No AI-specific meta directive found (not a defect -- this is not an established standard)
CATEGORY
Discovery & Crawl Access
ARC TEST ID
meta_robots_ai_directive_detected
COVERED BY
Remediation item R-01 — see the prioritised roadmap.
F-09 · MACHINE READABILITY & CONTENT STRUCTURE
LOW
FAIL
llms-full.txt is available
WHAT ARC FOUND
llms-full.txt not found or too short
CATEGORY
Machine Readability & Content Structure
ARC TEST ID
llms_full_txt_available
COVERED BY
Remediation item R-04 — see the prioritised roadmap.
F-10 · MACHINE READABILITY & CONTENT STRUCTURE
LOW
FAIL
llms.txt is available
WHAT ARC FOUND
llms.txt not found or not well-formed
WHAT SHOULD HAPPEN
Draft and publish an llms.txt summary.
CATEGORY
Machine Readability & Content Structure
ARC TEST ID
llms_txt_available
COVERED BY
Remediation item R-04 — see the prioritised roadmap.
F-11 · MACHINE READABILITY & CONTENT STRUCTURE
LOW
NEEDS ATTENTION
Markdown content negotiation or fallback is available
WHAT ARC FOUND
No Markdown content negotiation or /index.md fallback was found (not required).
CATEGORY
Machine Readability & Content Structure
ARC TEST ID
markdown_content_negotiation
COVERED BY
Remediation item R-04 — see the prioritised roadmap.
EXECUTED CHECKS WITH NO VERDICT — NOT PASSES
| CHECK |
CATEGORY |
RESULT |
WHY NO VERDICT |
|---|
| A2A agent card is valid a2a_agent_card_valid |
Agent Interfaces & Actionability |
NOT APPLICABLE |
No A2A agent card was found at /.well-known/agent-card.json. |
| Agent Skills are discoverable and well-formed agent_skills_discoverable |
Agent Interfaces & Actionability |
NOT VERIFIED |
No publicly documented, standard discovery location exists yet for an Agent Skills manifest on a public website; automating this would require inventing a non-standard convention, so it is always reported as Not verified pending a documented standard. |
| MCP endpoint or manifest is discoverable mcp_descriptor_discoverable |
Agent Interfaces & Actionability |
NOT VERIFIED |
No documented, single well-known path exists for MCP manifest discovery, and no linked/referenced MCP manifest was found on the homepage or in llms.txt -- this check cannot distinguish absence from an undiscoverable-by-convention manifest, so it is reported as Not verified rather than guessed. |
| OAuth Authorization Server Metadata is discoverable oauth_authorization_server_metadata |
Identity, Authentication & Agent Access |
NOT APPLICABLE |
No OAuth Authorization Server Metadata was found. |
| OAuth Protected Resource Metadata is discoverable oauth_protected_resource_metadata |
Identity, Authentication & Agent Access |
NOT APPLICABLE |
No OAuth Protected Resource Metadata was found. |
| Freshness/date information is present where material freshness_signals_present |
Machine Readability & Content Structure |
NOT APPLICABLE |
No date/freshness signal was found even though pricing/offer language is present. |
| No duplicate/canonical content ambiguity detected duplicate_canonical_ambiguity |
Machine Readability & Content Structure |
NOT VERIFIED |
Not evaluated at this scan depth; duplicate-content comparison requires the full multi-page sample. |
| Structured data types are relevant to the business schema_types_relevant_to_business |
Machine Readability & Content Structure |
NOT APPLICABLE |
No structured data present to assess relevance. |
Fictitious sample · made-up data
06
Category assessments
Category by category: what ARC evaluated, what it scored, and what it could not score. Counts are executed automated checks; operator-verified checks are reported separately.
Discovery & Crawl Access
SCORED
86/100
Evaluated 11 ARC checks in this category. 9 of 11 checks passed.
WHAT ARC TESTED
This category asks: Can AI systems find and access the information you want them to see?
ARC considered 11 automated checks in this category for this assessment: Homepage is fetchable; robots.txt is available; Site is not globally blocked; Named AI crawlers are not blocked; Redirect chain and final host are correct; XML sitemap is available; Sitemap is valid XML; Meta robots / X-Robots-Tag restrictions are documented; WAF/CDN challenge does not block plain HTTP access; Primary content does not require JavaScript execution; AI-specific meta directives detected.
WHAT ARC OBSERVED
Of 11 automated checks: 9 passed, 1 need attention, 1 failed, 0 not verified, 0 not applicable.
- Site is not globally blocked — FAIL: Global crawling is blocked
- AI-specific meta directives detected — NEEDS ATTENTION: No AI-specific meta directive found (not a defect -- this is not an established standard)
WHY THIS RESULT
The category score of 86/100 comes from 11 scored checks (9 passed, 1 need attention, 1 failed). Only pass, needs-attention and fail results carry points.
Failed checks mean the expected condition was not found in the evidence ARC recorded.
WHY IT MATTERS
- Site is not globally blocked: A blanket disallow tells every crawler, including AI agents, not to access any page on the site.
- AI-specific meta directives detected: Surfaces an explicit AI-training opt-out signal where a site owner has chosen to set one, without treating its absence as a defect.
NEXT STEP
- Publish or correct robots.txt and remove unintended global or AI-crawler blocks.
FINDING
Site is not globally blocked: Global crawling is blocked
FINDING
AI-specific meta directives detected: No AI-specific meta directive found (not a defect -- this is not an established standard)
Machine Readability & Content Structure
SCORED
50/100
Evaluated 11 ARC checks in this category. 5 of 11 checks passed.
WHAT ARC TESTED
This category asks: Can machines reliably read and interpret the important content on your site?
ARC considered 13 automated checks in this category for this assessment: llms.txt is available; llms-full.txt is available; Markdown content negotiation or fallback is available; Descriptive title is present; Meta description is present; Canonical URL is declared; Valid JSON-LD exists; Structured data types are relevant to the business; Service/Product/Offer/FAQ schema exists; Structured data is consistent with visible content; No duplicate/canonical content ambiguity detected; Freshness/date information is present where material; Agentic Web signals present.
WHAT ARC OBSERVED
Of 13 automated checks: 5 passed, 1 need attention, 4 failed, 1 not verified, 2 not applicable.
- Valid JSON-LD exists — FAIL: Types: none
- Service/Product/Offer/FAQ schema exists — FAIL: Service/Product/Offer/FAQ schema missing
- llms.txt is available — FAIL: llms.txt not found or not well-formed
- llms-full.txt is available — FAIL: llms-full.txt not found or too short
- Structured data types are relevant to the business — NOT APPLICABLE: No structured data present to assess relevance.
- Freshness/date information is present where material — NOT APPLICABLE: No date/freshness signal was found even though pricing/offer language is present.
WHY THIS RESULT
The category score of 50/100 comes from 10 scored checks (5 passed, 1 need attention, 4 failed). Only pass, needs-attention and fail results carry points.
Failed checks mean the expected condition was not found in the evidence ARC recorded.
3 checks were excluded from this score (2 not applicable, 1 not verified); they neither raised nor lowered it.
WHY IT MATTERS
- Valid JSON-LD exists: JSON-LD is a primary machine-readable format for structured facts. Structured data is valuable for search features and agent parsing, but its presence is not itself a guarantee of inclusion in Google AI Overviews or any specific AI product.
- Service/Product/Offer/FAQ schema exists: Lets an agent resolve exactly what is being sold and its attributes, not just that the page mentions a product.
- llms.txt is available: An emerging, optional convention giving AI agents a curated summary instead of raw HTML. Not a web standard — and Google has stated Search does not use llms.txt for its generative AI features — so absence is informational, not a critical failure.
- llms-full.txt is available: Where present, gives agents that consume the llms.txt convention a fuller reference than the summary file alone. Same non-standard, informational status as llms.txt.
NEXT STEP
- Add and validate schema.org JSON-LD covering business entity identity and core offers.
- Draft and publish an llms.txt summary.
FINDING
Valid JSON-LD exists: Types: none
FINDING
Service/Product/Offer/FAQ schema exists: Service/Product/Offer/FAQ schema missing
FINDING
llms.txt is available: llms.txt not found or not well-formed
Business Understanding & Answerability
SCORED
89/100
Evaluated 6 ARC checks in this category. 5 of 6 checks passed.
WHAT ARC TESTED
This category asks: Can AI correctly explain what your business offers and answer the questions customers actually ask?
ARC considered 6 automated checks in this category for this assessment: Content is substantial enough to quote; Company/entity name is clear; Entity schema exists; Contact identity is visible; About, policies, or proof signals exist; Key buyer questions are answerable from retrieved content.
WHAT ARC OBSERVED
Of 6 automated checks: 5 passed, 0 need attention, 1 failed, 0 not verified, 0 not applicable.
- Entity schema exists — FAIL: Organization/LocalBusiness entity schema missing
WHY THIS RESULT
The category score of 89/100 comes from 6 scored checks (5 passed, 0 need attention, 1 failed). Only pass, needs-attention and fail results carry points.
Failed checks mean the expected condition was not found in the evidence ARC recorded.
WHY IT MATTERS
- Entity schema exists: Gives an agent a structured, machine-verifiable identity record instead of an inferred one.
NEXT STEP
- Publish Organization/LocalBusiness schema matching authoritative business details.
FINDING
Entity schema exists: Organization/LocalBusiness entity schema missing
AI Visibility & Recommendation
PENDING MANUAL
—
No numeric score is reported for this category (pending manual). This is not a zero and not a failure.
WHAT ARC TESTED
This category asks: Does your business appear when customers use AI to research and compare providers?
ARC considered 0 automated checks and 1 operator-verified check in this category for this assessment: Do AI systems mention and recommend you?.
WHAT ARC OBSERVED
No automated check result was recorded for this category. 1 operator-verified check has no recorded result yet.
WHY THIS RESULT
No automated check in this category produced a scoreable result, and the category depends on operator-verified check with no recorded result yet. No numeric score is reported. This is not a zero and not a failure.
WHY IT MATTERS
- Do AI systems mention and recommend you?: Technical readiness does not guarantee an AI system actually surfaces or recommends the business — this is the direct, model-observed signal of whether it does.
NEXT STEP
Complete the pending operator verification (see "What is missing").
WHAT IS MISSING
- Do AI systems mention and recommend you?: awaiting operator verification. Benchmarks realistic non-branded buyer-intent prompts for this business's category against major AI answer engines (ChatGPT/OpenAI, Gemini, Claude, Grok, Perplexity where appropriate), recording brand mention rate, citation rate, prominence, category recommendation presence, and competitor share of voice. Performed manually by Hoshiro during paid-report fulfilment for this launch phase — not scanner-automated.
IF THIS LOOKS WRONG
If you believe the manual verification is complete or mis-scoped, contact Hoshiro Labs (reply to your report delivery email).
PENDING MANUAL
Do AI systems mention and recommend you? — operator-verified check.
Agent Interfaces & Actionability
SCORED
67/100
Evaluated 6 ARC checks in this category. 2 of 6 checks passed.
WHAT ARC TESTED
This category asks: Does your site expose reliable ways for AI agents to use its functions instead of guessing through the UI?
ARC considered 7 automated checks and 3 operator-verified checks in this category for this assessment: Form or reachable contact path exists; OpenAPI documentation is valid; API Catalog is discoverable; MCP endpoint or manifest is discoverable; Agent Skills are discoverable and well-formed; A2A agent card is valid; Public transaction interface is documented; Your MCP server actually works; Your agent-to-agent service actually responds correctly; Your in-page AI tools actually work.
WHAT ARC OBSERVED
Of 7 automated checks: 2 passed, 0 need attention, 2 failed, 2 not verified, 1 not applicable. 3 operator-verified checks have no recorded result yet.
- OpenAPI documentation is valid — FAIL: No parseable OpenAPI/Swagger document was found.
- API Catalog is discoverable — FAIL: No API Catalog was found.
- A2A agent card is valid — NOT APPLICABLE: No A2A agent card was found at /.well-known/agent-card.json.
- MCP endpoint or manifest is discoverable — NOT VERIFIED: No documented, single well-known path exists for MCP manifest discovery, and no linked/referenced MCP manifest was found on the homepage or in llms.txt -- this check cannot distinguish absence from an undiscoverable-by-convention manifest, so it is reported as Not verified rather than guessed.
WHY THIS RESULT
The category score of 67/100 comes from 4 scored checks (2 passed, 0 need attention, 2 failed). Only pass, needs-attention and fail results carry points.
Failed checks mean the expected condition was not found in the evidence ARC recorded.
3 checks were excluded from this score (1 not applicable, 2 not verified); they neither raised nor lowered it.
WHY IT MATTERS
- OpenAPI documentation is valid: Only relevant where a real commercial API use case exists; this scanner never invents an API as a requirement.
- API Catalog is discoverable: A standard catalog location lets an agent discover all of a site's APIs without guessing paths.
- Your MCP server actually works: A published MCP descriptor that fails a real handshake gives agents nothing — real behaviour matters more than file presence.
- Your agent-to-agent service actually responds correctly: A valid-looking agent card does not guarantee the underlying service behaves correctly under a real task exchange.
- Your in-page AI tools actually work: A page that merely references WebMCP without correctly registered, usable tools gives an agent nothing more than a descriptor would.
NEXT STEP
No remediation is required for the checks evaluated in this category.
FINDING
OpenAPI documentation is valid: No parseable OpenAPI/Swagger document was found.
FINDING
API Catalog is discoverable: No API Catalog was found.
PENDING MANUAL
Your MCP server actually works — operator-verified check.
PENDING MANUAL
Your agent-to-agent service actually responds correctly — operator-verified check.
PENDING MANUAL
Your in-page AI tools actually work — operator-verified check.
Identity, Authentication & Agent Access
NOT APPLICABLE
—
No numeric score is reported for this category (not applicable). This is not a zero and not a failure.
WHAT ARC TESTED
This category asks: Can legitimate agents identify themselves and securely access the capabilities they are authorised to use?
ARC considered 2 automated checks and 1 operator-verified check in this category for this assessment: OAuth Authorization Server Metadata is discoverable; OAuth Protected Resource Metadata is discoverable; Agent sign-in actually works as documented.
- OAuth Authorization Server Metadata is discoverable — /.well-known/oauth-authorization-server is fetchable and validates per RFC 8414. Whether it is expected is decided from the business surface and declared agent policy, not from the file's presence: Not applicable when the established surface is public/enquiry-only or the business declared agents may not act; NOT VERIFIED when the surface or intent cannot be established; evaluated when an authenticated/agent-facing surface or declared agent intent exists.
- OAuth Protected Resource Metadata is discoverable — /.well-known/oauth-protected-resource is fetchable and validates per RFC 9728. Applicability follows the business surface and declared intent (see oauth_authorization_server_metadata). A FAIL requires proof: an agent-facing endpoint that demonstrably answered HTTP 401 while advertising no discovery path.
WHAT ARC OBSERVED
Of 2 automated checks: 0 passed, 0 need attention, 0 failed, 0 not verified, 2 not applicable. 1 operator-verified check has no recorded result yet.
- OAuth Authorization Server Metadata is discoverable — NOT APPLICABLE: No OAuth Authorization Server Metadata was found.
- OAuth Protected Resource Metadata is discoverable — NOT APPLICABLE: No OAuth Protected Resource Metadata was found.
WHY THIS RESULT
Every automated check that ran in this category (2 of 2) was not applicable to this site, so no check carried points. A category with nothing scoreable is shown as NOT APPLICABLE with no numeric score. It is not a zero and not a failure, and the excluded checks did not lower any category result.
For these checks ARC probed for the surface they assess and did not establish that it exists, so they were excluded from scoring rather than failed.
1 operator-verified check in this category remains pending (see below).
WHY IT MATTERS
- OAuth Authorization Server Metadata is discoverable: Standard OAuth metadata lets an agent discover authorization/token endpoints and scopes without out-of-band instructions.
- OAuth Protected Resource Metadata is discoverable: Lets an agent locate the authorization server responsible for a protected resource without prior knowledge.
- Agent sign-in actually works as documented: Advertised auth metadata that does not match real service behaviour breaks every agent that trusts it.
NEXT STEP
No action is required unless you operate the capability described under "When this would apply".
WHEN THIS WOULD APPLY
- OAuth Authorization Server Metadata is discoverable; OAuth Protected Resource Metadata is discoverable: Applies only when the business surface ARC established includes an authenticated or consequential action surface and the business intends agents to act on it (declared), or an agent-facing interface is published or demonstrably requires authentication. A public or enquiry-only site, or one that declares agents must not act, is outside scope.
- Agent sign-in actually works as documented: Applies only when the site exposes a delegated or protected agent-access surface, for example an OAuth-protected API, an MCP or A2A service with protected tools, customer accounts that agents can act within, or another authenticated agent-access surface.
IF THIS LOOKS WRONG
If you operate an authenticated API, an MCP or A2A service with protected tools, customer accounts, or another agent-access surface that ARC did not detect or an authenticated or consequential action surface, or an agent-facing interface, that ARC did not detect or that the business intends agents to use, send Hoshiro Labs the relevant URL or endpoint (for example its well-known metadata location or API base URL) and request a reassessment of this category. Contact Hoshiro Labs (reply to your report delivery email).
PENDING MANUAL
Agent sign-in actually works as documented — operator-verified check.
Conversion & Agentic Commerce
SCORED
88/100
Evaluated 3 ARC checks in this category. 2 of 3 checks passed.
WHAT ARC TESTED
This category asks: Can an AI agent progress toward a real business outcome such as an enquiry, booking, signup or purchase?
ARC considered 3 automated checks and 1 operator-verified check in this category for this assessment: Clear conversion action is visible; Pricing or offer context is visible; Emerging agentic-commerce protocol signals; An AI agent can actually complete your primary action.
WHAT ARC OBSERVED
Of 3 automated checks: 2 passed, 1 need attention, 0 failed, 0 not verified, 0 not applicable. 1 operator-verified check has no recorded result yet.
- Pricing or offer context is visible — NEEDS ATTENTION: No public pricing/offer language detected (may be an intentional sales-led model, not a defect)
WHY THIS RESULT
The category score of 88/100 comes from 3 scored checks (2 passed, 1 need attention, 0 failed). Only pass, needs-attention and fail results carry points.
Needs-attention checks mean something relevant is present but incomplete, informational or constrained.
WHY IT MATTERS
- Pricing or offer context is visible: Helps an agent compare options quickly; a 'contact sales' model is a legitimate choice, not a defect by itself.
- An AI agent can actually complete your primary action: This is the direct test of whether an agent can actually get a customer to the outcome the business needs, not just whether the page looks navigable.
NEXT STEP
No remediation is required for the checks evaluated in this category.
FINDING
Pricing or offer context is visible: No public pricing/offer language detected (may be an intentional sales-led model, not a defect)
PENDING MANUAL
An AI agent can actually complete your primary action — operator-verified check.
Fictitious sample · made-up data
07
Manual verification status
These checks are performed by a Hoshiro operator during paid-report fulfilment, never by the automated scanner. A check with no recorded result is NOT VERIFIED — not a pass.
Do AI systems mention and recommend you?
NOT VERIFIED · PENDING
AI Visibility & Recommendation assessment · Technical readiness does not guarantee an AI system actually surfaces or recommends the business — this is the direct, model-observed signal of whether it does.
NOT VERIFIED — assessed manually, pending fulfilment. No result has been recorded for this check yet; this is never scored or implied as a pass. ai_visibility_brand_mention_citation
- What must be verified
- Benchmarks realistic non-branded buyer-intent prompts for this business's category against major AI answer engines (ChatGPT/OpenAI, Gemini, Claude, Grok, Perplexity where appropriate), recording brand mention rate, citation rate, prominence, category recommendation presence, and competitor share of voice. Performed manually by Hoshiro during paid-report fulfilment for this launch phase — not scanner-automated.
- Why ARC does not do this automatically
- This check is defined as operator-verified. The automated scanner never performs it.
- Credentials or customer cooperation
- The check definition does not state that credentials are required.
- Is the report incomplete for this capability?
- Yes. This capability is NOT VERIFIED in this report until a result is recorded. It is not scored, and it is neither a pass nor a failure.
- How to complete it
- Contact Hoshiro Labs (reply to your report delivery email) to arrange the verification. Where access is required, it must come from you.
- Why it matters
- Technical readiness does not guarantee an AI system actually surfaces or recommends the business — this is the direct, model-observed signal of whether it does.
Your MCP server actually works
NOT VERIFIED · PENDING
MCP handshake and tool quality (manual verification) · A published MCP descriptor that fails a real handshake gives agents nothing — real behaviour matters more than file presence.
NOT VERIFIED — assessed manually, pending fulfilment. No result has been recorded for this check yet; this is never scored or implied as a pass. mcp_real_handshake_verified
- What must be verified
- A real MCP connection is attempted (initialize + tools/list JSON-RPC over the Streamable HTTP transport) against the referenced manifest's endpoint, and returned tool names/descriptions/input schemas are checked for clarity and non-overlap. No tool is ever invoked.
- Why ARC does not do this automatically
- This report records the live interaction for this check as operator-verified, so it is completed by a Hoshiro operator rather than by the automated scan.
- Credentials or customer cooperation
- The check definition does not state that credentials are required.
- When it applies
- Applies only when the site exposes a delegated or protected agent-access surface, for example an OAuth-protected API, an MCP or A2A service with protected tools, customer accounts that agents can act within, or another authenticated agent-access surface.
- Relationship to the automated checks
- The automated checks for the same surface (A2A agent card is valid) were not applicable, so there may be nothing for an operator to verify unless you operate that surface.
- Is the report incomplete for this capability?
- Yes. This capability is NOT VERIFIED in this report until a result is recorded. It is not scored, and it is neither a pass nor a failure.
- How to complete it
- Contact Hoshiro Labs (reply to your report delivery email) to arrange the verification. Where access is required, it must come from you.
- Why it matters
- A published MCP descriptor that fails a real handshake gives agents nothing — real behaviour matters more than file presence.
Your agent-to-agent service actually responds correctly
NOT VERIFIED · PENDING
A2A behavioral verification (manual) · A valid-looking agent card does not guarantee the underlying service behaves correctly under a real task exchange.
NOT VERIFIED — assessed manually, pending fulfilment. No result has been recorded for this check yet; this is never scored or implied as a pass. a2a_behavioral_verification
- What must be verified
- Where a valid agent card's `url` is present, a real A2A `message/send` task exchange is attempted with a short, generic capability-discovery message (never a transactional one).
- Why ARC does not do this automatically
- This report records the live interaction for this check as operator-verified, so it is completed by a Hoshiro operator rather than by the automated scan.
- Credentials or customer cooperation
- The check definition does not state that credentials are required.
- When it applies
- Applies only when the site exposes a delegated or protected agent-access surface, for example an OAuth-protected API, an MCP or A2A service with protected tools, customer accounts that agents can act within, or another authenticated agent-access surface.
- Relationship to the automated checks
- The automated checks for the same surface (A2A agent card is valid) were not applicable, so there may be nothing for an operator to verify unless you operate that surface.
- Is the report incomplete for this capability?
- Yes. This capability is NOT VERIFIED in this report until a result is recorded. It is not scored, and it is neither a pass nor a failure.
- How to complete it
- Contact Hoshiro Labs (reply to your report delivery email) to arrange the verification. Where access is required, it must come from you.
- Why it matters
- A valid-looking agent card does not guarantee the underlying service behaves correctly under a real task exchange.
Your in-page AI tools actually work
NOT VERIFIED · PENDING
WebMCP tools actually work (manual) · A page that merely references WebMCP without correctly registered, usable tools gives an agent nothing more than a descriptor would.
NOT VERIFIED — assessed manually, pending fulfilment. No result has been recorded for this check yet; this is never scored or implied as a pass. webmcp_real_interaction_verified
- What must be verified
- Where WebMCP tool registration is detected on the page, real tool registration, naming, schema usability, and a safe task execution are verified manually during paid fulfilment.
- Why ARC does not do this automatically
- This check is defined as operator-verified. The automated scanner never performs it.
- Credentials or customer cooperation
- The check definition does not state that credentials are required.
- When it applies
- Applies only when the site exposes a delegated or protected agent-access surface, for example an OAuth-protected API, an MCP or A2A service with protected tools, customer accounts that agents can act within, or another authenticated agent-access surface.
- Relationship to the automated checks
- The automated checks for the same surface (A2A agent card is valid) were not applicable, so there may be nothing for an operator to verify unless you operate that surface.
- Is the report incomplete for this capability?
- Yes. This capability is NOT VERIFIED in this report until a result is recorded. It is not scored, and it is neither a pass nor a failure.
- How to complete it
- Contact Hoshiro Labs (reply to your report delivery email) to arrange the verification. Where access is required, it must come from you.
- Why it matters
- A page that merely references WebMCP without correctly registered, usable tools gives an agent nothing more than a descriptor would.
Agent sign-in actually works as documented
NOT VERIFIED · PENDING
Authenticated agent access verification (manual) · Advertised auth metadata that does not match real service behaviour breaks every agent that trusts it.
NOT VERIFIED — assessed manually, pending fulfilment. No result has been recorded for this check yet; this is never scored or implied as a pass. authenticated_agent_access_verification
- What must be verified
- Where OAuth/agent-login metadata is published, a real authenticated flow (PKCE S256 support, usable scopes, agreement between advertised metadata and the actual protected service) is verified manually during paid fulfilment. Never attempted automatically without credentials.
- Why ARC does not do this automatically
- This check is defined as operator-verified. The automated scanner never performs it.
- Credentials or customer cooperation
- Yes. The check definition states it is never attempted without credentials, so it needs access supplied by you.
- When it applies
- Applies only when the site exposes a delegated or protected agent-access surface, for example an OAuth-protected API, an MCP or A2A service with protected tools, customer accounts that agents can act within, or another authenticated agent-access surface.
- Relationship to the automated checks
- The automated checks for the same surface (OAuth Authorization Server Metadata is discoverable, OAuth Protected Resource Metadata is discoverable) were not applicable, so there may be nothing for an operator to verify unless you operate that surface.
- Is the report incomplete for this capability?
- Yes. This capability is NOT VERIFIED in this report until a result is recorded. It is not scored, and it is neither a pass nor a failure.
- How to complete it
- Contact Hoshiro Labs (reply to your report delivery email) to arrange the verification. Where access is required, it must come from you.
- Why it matters
- Advertised auth metadata that does not match real service behaviour breaks every agent that trusts it.
An AI agent can actually complete your primary action
NOT VERIFIED · PENDING
Agent task completion is verified (manual) · This is the direct test of whether an agent can actually get a customer to the outcome the business needs, not just whether the page looks navigable.
NOT VERIFIED — assessed manually, pending fulfilment. No result has been recorded for this check yet; this is never scored or implied as a pass. agent_task_completion_verified
- What must be verified
- A safe, non-destructive attempt to progress through the site's actual primary conversion action (enquiry/booking/signup/purchase) as an agent would, capturing where the task succeeds or breaks. Performed manually during paid fulfilment; never submits a destructive or chargeable action during an ordinary scan.
- Why ARC does not do this automatically
- This check is defined as operator-verified. The automated scanner never performs it.
- Credentials or customer cooperation
- Access may be needed if the conversion flow is gated; if it is, the result is recorded as Blocked · Pending access rather than as a pass or a failure.
- Is the report incomplete for this capability?
- Yes. This capability is NOT VERIFIED in this report until a result is recorded. It is not scored, and it is neither a pass nor a failure.
- How to complete it
- Contact Hoshiro Labs (reply to your report delivery email) to arrange the verification. Where access is required, it must come from you.
- Why it matters
- This is the direct test of whether an agent can actually get a customer to the outcome the business needs, not just whether the page looks navigable.
Fictitious sample · made-up data
08
Prioritised remediation roadmap
Items are ordered by priority. Each states the problem, the outcome Hoshiro would deliver, the ARC tests it is verified against, and its commercial status. Implementation method is deliberately out of scope for this document.
R-01
CRITICAL
AI Crawler Access & Discoverability
USD 1,800.00
PROBLEM
A blanket disallow tells every crawler, including AI agents, not to access any page on the site. Surfaces an explicit AI-training opt-out signal where a site owner has chosen to set one, without treating its absence as a defect.
DESIRED OUTCOME
Publish or correct robots.txt and XML sitemap, remove unintended global or AI-crawler blocks, and declare a canonical URL for the primary site templates. Verified after deployment against the same evaluation criteria used in this audit.
HOSHIRO SCOPE
Correct technical controls that prevent or confuse AI/search crawlers when accessing the website, including relevant crawler policies and discovery configuration identified in the audit.
RELATED ARC TESTS
meta_robots_ai_directive_detected, site_not_globally_blocked
CUSTOMER PREREQUISITES
CMS or hosting access · DNS/Cloudflare access if applicable
ESTIMATED DURATION
1-2 weeks
COMMERCIAL STATUS
Indicative price available
R-02
HIGH
Structured Data & Entity Implementation
USD 2,400.00
PROBLEM
JSON-LD is a primary machine-readable format for structured facts. Structured data is valuable for search features and agent parsing, but its presence is not itself a guarantee of inclusion in any specific AI product. Lets an agent resolve exactly what is being sold and its attributes, and gives it a structured, machine-verifiable identity record instead of an inferred one.
DESIRED OUTCOME
Add and validate schema.org JSON-LD covering business entity identity and the site's core services/products/offers, matching what is genuinely offered. Verified after deployment against the same evaluation criteria used in this audit.
HOSHIRO SCOPE
Improve machine-readable descriptions of the company, products, services and key content so AI systems can more reliably understand the organisation and its offering.
RELATED ARC TESTS
entity_schema_exists, service_product_offer_faq_schema, valid_json_ld_exists
CUSTOMER PREREQUISITES
CMS or repository access · Confirmation of authoritative business details
ESTIMATED DURATION
2-3 weeks
COMPLEXITY
Platform-Assisted
COMMERCIAL STATUS
Indicative price available
R-03
LOW
AI Agent Action & Conversion Readiness
Custom Scope Review Required
PROBLEM
Helps an agent compare options quickly; a 'contact sales' model is a legitimate choice, not a defect by itself.
DESIRED OUTCOME
Add or clarify primary call-to-action links/buttons and ensure a reliable form or contact path exists with clear success/error states. Verified after deployment against the same evaluation criteria used in this audit.
HOSHIRO SCOPE
Improve customer pathways that AI-assisted users or agents need to move from information to meaningful action, such as contacting, requesting a quote, booking, buying or accessing a service.
RELATED ARC TESTS
pricing_offer_context_visible
CUSTOMER PREREQUISITES
CMS access · Staging environment if available
ESTIMATED DURATION
To be scoped
COMPLEXITY
Custom Scope Review Required
COMMERCIAL STATUS
Custom Scope Review Required
R-04
LOW
Agent-Friendly Documentation & Discovery
Custom Scope Review Required
PROBLEM
An emerging, optional convention giving AI agents a curated summary instead of raw HTML. Not a web standard, so absence is informational, not a critical failure. Some agent tooling prefers Markdown over HTML for lower-noise parsing; support is emerging, not required.
DESIRED OUTCOME
Draft and publish an llms.txt summary and/or scope a minimal documented interface for genuinely agent-actionable transactions. Verified after deployment against the same evaluation criteria used in this audit.
HOSHIRO SCOPE
Improve public documentation and discovery resources that help automated systems understand the site, services and available interfaces, including appropriate emerging Agentic Web conventions.
RELATED ARC TESTS
api_catalog_available, llms_full_txt_available, llms_txt_available, markdown_content_negotiation, openapi_documentation_valid
CUSTOMER PREREQUISITES
CMS access · Technical contact for API scope confirmation
ESTIMATED DURATION
To be scoped
COMPLEXITY
Custom Scope Review Required
COMMERCIAL STATUS
Custom Scope Review Required
Fictitious sample · made-up data
09
Commercial & scope summary
| REMEDIATION ITEM |
PRIORITY |
DURATION |
INDICATIVE PRICE |
|---|
| AI Crawler Access & Discoverability R-01 |
CRITICAL |
1-2 weeks |
USD 1,800.00 |
| Structured Data & Entity Implementation R-02 |
HIGH |
2-3 weeks |
USD 2,400.00 |
| AI Agent Action & Conversion Readiness R-03 |
LOW |
To be scoped |
Custom Scope Review Required |
| Agent-Friendly Documentation & Discovery R-04 |
LOW |
To be scoped |
Custom Scope Review Required |
| 4 items — subtotal covers 2 priceable items only |
USD 4,200.00 |
EXCLUDED FROM THE SUBTOTAL
The subtotal covers priceable items only. 2 items are excluded pending scope review: AI Agent Action & Conversion Readiness; Agent-Friendly Documentation & Discovery. Custom-scope work is never implied to cost zero.
TIMELINE
Indicative timeline covers scoped items only; remaining items subject to scope confirmation. 2 of 4 items are not yet scoped, so no combined delivery estimate is stated.
These are indicative remediation prices based on the findings in this audit, not an accepted commercial agreement. Final selected scope will be confirmed in a Statement of Work / Work Order. Estimates assume the stated access and dependencies for each item; a major scope change may require repricing.
Fictitious sample · made-up data
10
What ARC tested — coverage summary
Partial coverage — one or more pages in the sample could not be reliably assessed; treat findings as directional for those pages. Fetched and reviewed on 9 of 12 sampled pages. Detected platform: Unknown / Manual Assessment Required.
EXECUTED TEST COVERAGE BY CATEGORY
| CATEGORY |
EXECUTED |
PASS |
NEEDS ATTENTION |
FAIL |
NOT VERIFIED |
N/A |
STATE |
| Discovery & Crawl Access |
11 |
9 |
1 |
1 |
0 |
0 |
SCORED |
| Machine Readability & Content Structure |
13 |
5 |
1 |
4 |
1 |
2 |
SCORED |
| Business Understanding & Answerability |
6 |
5 |
0 |
1 |
0 |
0 |
SCORED |
| AI Visibility & Recommendation |
0 |
0 |
0 |
0 |
0 |
0 |
PENDING MANUAL |
| Agent Interfaces & Actionability |
7 |
2 |
0 |
2 |
2 |
1 |
SCORED |
| Identity, Authentication & Agent Access |
2 |
0 |
0 |
0 |
0 |
2 |
NOT APPLICABLE |
| Conversion & Agentic Commerce |
3 |
2 |
1 |
0 |
0 |
0 |
SCORED |
| Total |
42 |
23 |
3 |
8 |
3 |
5 |
|
OPERATOR-VERIFIED CHECKS · REPORTED SEPARATELY, NEVER SCORED WHILE PENDING
| CHECK |
STATE |
ARC TEST ID |
|---|
| Do AI systems mention and recommend you? |
NOT VERIFIED · PENDING |
ai_visibility_brand_mention_citation |
| Your MCP server actually works |
NOT VERIFIED · PENDING |
mcp_real_handshake_verified |
| Your agent-to-agent service actually responds correctly |
NOT VERIFIED · PENDING |
a2a_behavioral_verification |
| Your in-page AI tools actually work |
NOT VERIFIED · PENDING |
webmcp_real_interaction_verified |
| Agent sign-in actually works as documented |
NOT VERIFIED · PENDING |
authenticated_agent_access_verification |
| An AI agent can actually complete your primary action |
NOT VERIFIED · PENDING |
agent_task_completion_verified |
PAGE-LEVEL EVIDENCE, CONSOLIDATED
| PAGES |
OBSERVATION |
|---|
| 7 of 12 |
Fetched and reviewed for representative-page accessibility. |
| 3 of 12 |
Not reachable during this audit (status: 429). |
| 1 of 12 |
Primary scanned page; full ARC check set applied here. |
| 1 of 12 |
Reachable, but visible content was too sparse to meaningfully evaluate (may be client-rendered). |
Full per-URL evidence is in appendix B; it is not repeated in the narrative above.
Fictitious sample · made-up data
A
Detailed test manifest
Every automated check executed in this assessment, with its stable ARC test ID, result and recorded evidence. 42 checks, registry v2.0.0.
DISCOVERY & CRAWL ACCESS · 11 CHECKS
| CHECK |
RESULT |
EVIDENCE RECORDED BY ARC |
|---|
| Homepage is fetchable homepage_fetchable |
PASS |
HTTP status: 200- What it tests
- The homepage returns a successful (2xx-3xx) HTTP response.
- Standard
- HTTP/1.1 (RFC 9110)
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. HTTP status: 200.
- Scoring
- Included in scoring.
|
| robots.txt is available robots_txt_available |
PASS |
HTTP status: 200- What it tests
- /robots.txt returns a successful HTTP response.
- Standard
- RFC 9309
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. HTTP status: 200.
- Scoring
- Included in scoring.
|
| Site is not globally blocked site_not_globally_blocked |
FAIL |
Global crawling is blocked Next: Publish or correct robots.txt and remove unintended global or AI-crawler blocks. - In plain terms
- Your site isn't hidden from every crawler
- What it tests
- robots.txt does not contain a blanket Disallow: / for User-agent: *.
- Standard
- RFC 9309
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why this result
- Expected: robots.txt does not contain a blanket Disallow: / for User-agent: *. Found: Global crawling is blocked. The expected condition was not present in the evidence ARC recorded, so the check failed.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- A blanket disallow tells every crawler, including AI agents, not to access any page on the site.
- Recommendation
- Publish or correct robots.txt and remove unintended global or AI-crawler blocks.
- Scoring
- Included in scoring.
|
| Named AI crawlers are not blocked named_ai_crawlers_not_blocked |
PASS |
No explicit AI crawler block detected- What it tests
- robots.txt does not disallow any of the named AI crawlers this check recognizes (GPTBot, ClaudeBot, Google-Extended, CCBot, PerplexityBot, Amazonbot, Applebot-Extended).
- Standard
- RFC 9309 (named user-agent tokens)
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. No explicit AI crawler block detected.
- Scoring
- Included in scoring.
|
| Redirect chain and final host are correct redirect_chain_resolves_cleanly |
PASS |
0 redirect hop(s) to final host.- What it tests
- The homepage URL resolves in a small number of redirect hops to a single stable final host, with no redirect loop.
- Standard
- HTTP/1.1 (RFC 9110) redirection status codes
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. 0 redirect hop(s) to final host.
- Scoring
- Included in scoring.
|
| XML sitemap is available xml_sitemap_available |
PASS |
HTTP status: 200- What it tests
- /sitemap.xml (or the Sitemap: URL declared in robots.txt) returns a successful response.
- Standard
- sitemaps.org Sitemap protocol
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. HTTP status: 200.
- Scoring
- Included in scoring.
|
| Sitemap is valid XML xml_sitemap_valid |
PASS |
Sitemap parses as a well-formed sitemap document.- What it tests
- The fetched sitemap parses as well-formed XML matching the sitemap namespace, with at least one <url> entry.
- Standard
- sitemaps.org Sitemap protocol
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Sitemap parses as a well-formed sitemap document.
- Scoring
- Included in scoring.
|
| Meta robots / X-Robots-Tag restrictions are documented meta_robots_restrictions_documented |
PASS |
Meta robots/X-Robots-Tag directives found: 'none'.- What it tests
- Records any noindex, nofollow, nosnippet, or X-Robots-Tag directive found on the homepage, and whether it blocks AI-search visibility.
- Standard
- robots meta tag / X-Robots-Tag (Google/Bing convention)
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Meta robots/X-Robots-Tag directives found: 'none'.
- Scoring
- Included in scoring.
|
| WAF/CDN challenge does not block plain HTTP access waf_challenge_blocks_legitimate_access |
PASS |
Homepage response does not match known bot-challenge/interstitial markers (heuristic check).- What it tests
- The homepage response is not a bot-challenge/interstitial page (e.g. a JS-challenge or CAPTCHA wall) to a plain, well-identified HTTP request.
- Standard
- none — heuristic on response markers
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Homepage response does not match known bot-challenge/interstitial markers (heuristic check).
- Scoring
- Included in scoring.
|
| Primary content does not require JavaScript execution content_available_without_js |
PASS |
Static HTML already carries 3958 visible words; no JS-dependency signal detected.- What it tests
- Compares statically-fetched homepage content against a headless-render pass (where available); flags when meaningful content only appears after JS execution.
- Standard
- none — comparative heuristic
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Static HTML already carries 3958 visible words; no JS-dependency signal detected.
- Scoring
- Included in scoring.
|
| AI-specific meta directives detected meta_robots_ai_directive_detected |
NEEDS ATTENTION |
No AI-specific meta directive found (not a defect -- this is not an established standard)- In plain terms
- Explicit AI-training opt-out (if set)
- What it tests
- Records whether a non-standard meta directive such as noai/noimageai is present. Informational only (scoring_weight 0): it moves no score in either direction, since this is not an established standard.
- Standard
- none — non-standard meta convention
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why this result
- Something relevant is present but incomplete, informational or constrained: No AI-specific meta directive found (not a defect -- this is not an established standard). It earns partial credit and is reported as needing attention rather than as a failure.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- Surfaces an explicit AI-training opt-out signal where a site owner has chosen to set one, without treating its absence as a defect.
- Scoring
- Included in scoring.
|
MACHINE READABILITY & CONTENT STRUCTURE · 13 CHECKS
| CHECK |
RESULT |
EVIDENCE RECORDED BY ARC |
|---|
| llms.txt is available llms_txt_available |
FAIL |
llms.txt not found or not well-formed Next: Draft and publish an llms.txt summary. - In plain terms
- You publish an llms.txt summary
- What it tests
- /llms.txt is fetchable and looks like the llms.txt convention (Markdown with at least one heading).
- Standard
- llms.txt convention (non-standard, community-driven)
- How ARC checks it
- Checks that the expected file is served
- Why this result
- Expected: /llms.txt is fetchable and looks like the llms.txt convention (Markdown with at least one heading). Found: llms.txt not found or not well-formed. The expected condition was not present in the evidence ARC recorded, so the check failed.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- An emerging, optional convention giving AI agents a curated summary instead of raw HTML. Not a web standard — and Google has stated Search does not use llms.txt for its generative AI features — so absence is informational, not a critical failure.
- Recommendation
- Draft and publish an llms.txt summary.
- Scoring
- Included in scoring.
|
| llms-full.txt is available llms_full_txt_available |
FAIL |
llms-full.txt not found or too short- In plain terms
- You publish an expanded llms-full.txt
- What it tests
- /llms-full.txt is fetchable and non-trivial in length.
- Standard
- llms.txt convention (non-standard, community-driven)
- How ARC checks it
- Checks that the expected file is served
- Why this result
- Expected: /llms-full.txt is fetchable and non-trivial in length. Found: llms-full.txt not found or too short. The expected condition was not present in the evidence ARC recorded, so the check failed.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- Where present, gives agents that consume the llms.txt convention a fuller reference than the summary file alone. Same non-standard, informational status as llms.txt.
- Scoring
- Included in scoring.
|
| Markdown content negotiation or fallback is available markdown_content_negotiation |
NEEDS ATTENTION |
No Markdown content negotiation or /index.md fallback was found (not required).- In plain terms
- You offer a Markdown version of your pages
- What it tests
- Requesting the homepage with Accept: text/markdown returns Markdown, or an /index.md fallback exists.
- Standard
- HTTP content negotiation (RFC 9110 Accept header); no fixed community standard for the Markdown convention itself
- How ARC checks it
- Requests the resource and reads the HTTP response
- Why this result
- Something relevant is present but incomplete, informational or constrained: No Markdown content negotiation or /index.md fallback was found (not required). It earns partial credit and is reported as needing attention rather than as a failure.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- Some agent tooling prefers Markdown over HTML for lower-noise parsing; support is emerging, not required.
- Scoring
- Included in scoring.
|
| Descriptive title is present descriptive_title_present |
PASS |
Title present and descriptive- What it tests
- The homepage's <title> element is at least 10 characters long.
- Standard
- HTML Living Standard <title>
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Title present and descriptive.
- Scoring
- Included in scoring.
|
| Meta description is present meta_description_present |
PASS |
Description metadata found- What it tests
- A meta or og:description tag with non-empty content is present.
- Standard
- HTML meta description / Open Graph og:description
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Description metadata found.
- Scoring
- Included in scoring.
|
| Canonical URL is declared canonical_url_tag_present |
PASS |
Canonical URL declared- What it tests
- The homepage declares a <link rel="canonical"> pointing to an absolute URL.
- Standard
- HTML <link rel="canonical">
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Canonical URL declared.
- Scoring
- Included in scoring.
|
| Valid JSON-LD exists valid_json_ld_exists |
FAIL |
Types: none Next: Add and validate schema.org JSON-LD covering business entity identity and core offers. - In plain terms
- Your site has machine-readable structured data
- What it tests
- At least one schema.org @type is present in a JSON-LD block on the homepage (including client-rendered content where a render fallback succeeded).
- Standard
- schema.org / JSON-LD (W3C)
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why this result
- Expected: At least one schema.org @type is present in a JSON-LD block on the homepage (including client-rendered content where a render fallback succeeded). Found: Types: none. The expected condition was not present in the evidence ARC recorded, so the check failed.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- JSON-LD is a primary machine-readable format for structured facts. Structured data is valuable for search features and agent parsing, but its presence is not itself a guarantee of inclusion in Google AI Overviews or any specific AI product.
- Recommendation
- Add and validate schema.org JSON-LD covering business entity identity and core offers.
- Scoring
- Included in scoring.
|
| Structured data types are relevant to the business schema_types_relevant_to_business |
NOT APPLICABLE |
No structured data present to assess relevance.- In plain terms
- Your structured data matches your business type
- What it tests
- Where an entity/commercial schema type is present, checks it against an expected type family (e.g. LocalBusiness for a storefront, Organization generically) rather than crediting any JSON-LD at all.
- Standard
- schema.org type hierarchy
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why ARC treated it as outside scope
- The condition that makes this check applicable was not met for this site. ARC recorded: No structured data present to assess relevance.
- What makes it applicable
- Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- Generic or mismatched schema types (e.g. only WebSite/WebPage boilerplate) give agents little more than they'd get from prose.
- Scoring
- Not included in scoring: not applicable checks carry no points and do not affect the category score.
|
| Service/Product/Offer/FAQ schema exists service_product_offer_faq_schema |
FAIL |
Service/Product/Offer/FAQ schema missing- In plain terms
- Your offerings are marked up in structured data
- What it tests
- JSON-LD includes at least one of Service, Product, Offer, or FAQPage.
- Standard
- schema.org Service/Product/Offer/FAQPage
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why this result
- Expected: JSON-LD includes at least one of Service, Product, Offer, or FAQPage. Found: Service/Product/Offer/FAQ schema missing. The expected condition was not present in the evidence ARC recorded, so the check failed.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- Lets an agent resolve exactly what is being sold and its attributes, not just that the page mentions a product.
- Scoring
- Included in scoring.
|
| Structured data is consistent with visible content structured_data_matches_visible_content |
PASS |
No comparable structured-vs-visible fact (e.g. price) was found to cross-check.- What it tests
- Where both JSON-LD and visible text state a comparable fact (e.g. a price), checks they do not directly contradict each other.
- Standard
- schema.org / JSON-LD (W3C) consistency heuristic
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. No comparable structured-vs-visible fact (e.g. price) was found to cross-check.
- Scoring
- Included in scoring.
|
| No duplicate/canonical content ambiguity detected duplicate_canonical_ambiguity |
NOT VERIFIED |
Not evaluated at this scan depth; duplicate-content comparison requires the full multi-page sample.- In plain terms
- Your site isn't creating duplicate-content confusion
- What it tests
- Flags when sampled pages appear to duplicate the homepage content without a canonical tag resolving the ambiguity.
- Standard
- HTML <link rel="canonical"> consistency heuristic
- How ARC checks it
- Fetches the content and parses it against the expected structure
- What ARC attempted
- ARC ran this check as defined (Flags when sampled pages appear to duplicate the homepage content without a canonical tag resolving the ambiguity.) and recorded: Not evaluated at this scan depth; duplicate-content comparison requires the full multi-page sample.
- Why it could not be verified
- Live or deeper verification not performed in this assessment.
- What would resolve it
- Running that deeper step, or recording it through operator verification, would resolve this.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- Ambiguous duplicate content makes it harder for an agent to know which version to cite or trust.
- Scoring
- Not included in scoring: not verified checks carry no points. NOT VERIFIED means ARC could not reach a conclusion; it is not a failure.
|
| Freshness/date information is present where material freshness_signals_present |
NOT APPLICABLE |
No date/freshness signal was found even though pricing/offer language is present.- In plain terms
- Time-sensitive content shows a date
- What it tests
- Where a datePublished/dateModified schema field or a visible last-updated date is expected (e.g. pricing, offers), checks one is present.
- Standard
- schema.org dateModified/datePublished
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why ARC treated it as outside scope
- The condition that makes this check applicable was not met for this site. ARC recorded: No date/freshness signal was found even though pricing/offer language is present.
- What makes it applicable
- Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- An agent citing stale pricing or availability as current misleads the buyer it's assisting.
- Scoring
- Not included in scoring: not applicable checks carry no points and do not affect the category score.
|
| Agentic Web signals present agentic_web_signals_present |
PASS |
llms.txt, a machine-actionable API, or a referenced MCP manifest was found- What it tests
- Detects any publicly linked llms.txt, OpenAPI spec, or MCP-manifest-looking resource as a bundle of emerging-convention signals. Informational only (scoring_weight 0) as of 2.2.0: a count of conventions is not readiness, so presence adds nothing and absence costs nothing. The individual checks (llms.txt, OpenAPI, MCP) remain.
- Standard
- none — composite of llms.txt/OpenAPI/MCP descriptor presence
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. llms.txt, a machine-actionable API, or a referenced MCP manifest was found.
- Scoring
- Included in scoring.
|
BUSINESS UNDERSTANDING & ANSWERABILITY · 6 CHECKS
| CHECK |
RESULT |
EVIDENCE RECORDED BY ARC |
|---|
| Content is substantial enough to quote content_substantial_enough_to_quote |
PASS |
Visible word count: 3958- What it tests
- The homepage's visible text is at least 150 words (partial credit at 50+).
- Standard
- none — word-count heuristic
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Visible word count: 3958.
- Scoring
- Included in scoring.
|
| Company/entity name is clear entity_name_clear |
PASS |
Title provides entity context- What it tests
- The homepage has a non-empty <title>. This checks only for a page-title identity signal, not a verified business-registry identity.
- Standard
- HTML <title>
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Title provides entity context.
- Scoring
- Included in scoring.
|
| Entity schema exists entity_schema_exists |
FAIL |
Organization/LocalBusiness entity schema missing Next: Publish Organization/LocalBusiness schema matching authoritative business details. - In plain terms
- Your business identity is machine-readable
- What it tests
- JSON-LD includes an Organization, LocalBusiness, Corporation, Person, or ProfessionalService type.
- Standard
- schema.org Organization/LocalBusiness
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why this result
- Expected: JSON-LD includes an Organization, LocalBusiness, Corporation, Person, or ProfessionalService type. Found: Organization/LocalBusiness entity schema missing. The expected condition was not present in the evidence ARC recorded, so the check failed.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- Gives an agent a structured, machine-verifiable identity record instead of an inferred one.
- Recommendation
- Publish Organization/LocalBusiness schema matching authoritative business details.
- Scoring
- Included in scoring.
|
| Contact identity is visible contact_identity_visible |
PASS |
Contact signal detected- What it tests
- The homepage's text or links contain a contact/email/phone/address signal word.
- Standard
- none — text heuristic
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Contact signal detected.
- Scoring
- Included in scoring.
|
| About, policies, or proof signals exist about_policies_proof_signals |
PASS |
Trust/proof language detected- What it tests
- The homepage's text contains about/privacy/terms/testimonial/case-study/review language.
- Standard
- none — text heuristic
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Trust/proof language detected.
- Scoring
- Included in scoring.
|
| Key buyer questions are answerable from retrieved content business_answerability_extractive_coverage |
PASS |
Extractable support found for 6/9 buyer-question categories: cancellation, contact, cost, how_to_buy, terms, where. This measures whether information is present, clear and extractable from retrieved content -- it is not a live test of whether an AI answer engine actually answers correctly; that is assessed separately during managed paid fulfilment.- What it tests
- For each applicable buyer question (what/who/where/cost/how-to-buy/terms/cancellation/contact/differentiation), checks whether retrieved page content contains extractable supporting text, and records the answer, evidence URL, and confidence. Never fabricates an answer when no supporting text was found. This measures whether information is present, clear, and extractable — it is not a live test of whether an AI answer engine actually answers correctly; that is assessed separately during managed paid fulfilment.
- Standard
- none — rule-based extractive analysis over retrieved evidence
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Extractable support found for 6/9 buyer-question categories: cancellation, contact, cost, how_to_buy, terms, where. This measures whether information is present, clear and extractable from retrieved content -- it is not a live test of whether an AI answer engine actually answers correctly; that is assessed separately during managed paid fulfilment.
- Scoring
- Included in scoring.
|
AGENT INTERFACES & ACTIONABILITY · 7 CHECKS
| CHECK |
RESULT |
EVIDENCE RECORDED BY ARC |
|---|
| Form or reachable contact path exists form_or_contact_path_exists |
PASS |
Forms: 12; /contact: 200- What it tests
- The homepage contains an HTML <form>, or /contact returns a successful HTTP response.
- Standard
- HTML <form>
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Forms: 12; /contact: 200.
- Scoring
- Included in scoring.
|
| OpenAPI documentation is valid openapi_documentation_valid |
FAIL |
No parseable OpenAPI/Swagger document was found.- In plain terms
- Your API is documented in a machine-readable way
- What it tests
- A probed OpenAPI/Swagger document is served and actually parses as a valid OpenAPI/Swagger document (not merely a 200 response from an SPA catch-all). Not applicable when no API surface is indicated by independent evidence; NOT VERIFIED when a developer/API surface is indicated (e.g. a link to API docs) but no machine-readable description was found at the probed locations.
- Standard
- OpenAPI Specification
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why this result
- Expected: A probed OpenAPI/Swagger document is served and actually parses as a valid OpenAPI/Swagger document (not merely a 200 response from an SPA catch-all). Not applicable when no API surface is indicated by independent evidence; NOT VERIFIED when a developer/API surface is indicated (e.g. a link to API docs) but no machine-readable description was found at the probed locations. Found: No parseable OpenAPI/Swagger document was found. The expected condition was not present in the evidence ARC recorded, so the check failed.
- Applicability
- Applicable. Applies only when the site exposes a public API or developer surface, for example a documented API or an OpenAPI/Swagger description.
- Why it matters
- Only relevant where a real commercial API use case exists; this scanner never invents an API as a requirement.
- Scoring
- Included in scoring.
|
| API Catalog is discoverable api_catalog_available |
FAIL |
No API Catalog was found.- In plain terms
- Your APIs are listed in a standard catalog
- What it tests
- /.well-known/api-catalog is fetchable and parses per RFC 9727.
- Standard
- RFC 9727
- How ARC checks it
- Checks that the expected file is served
- Why this result
- Expected: /.well-known/api-catalog is fetchable and parses per RFC 9727. Found: No API Catalog was found. The expected condition was not present in the evidence ARC recorded, so the check failed.
- Applicability
- Applicable. Applies only when the site exposes a public API or developer surface, for example a documented API or an OpenAPI/Swagger description.
- Why it matters
- A standard catalog location lets an agent discover all of a site's APIs without guessing paths.
- Scoring
- Included in scoring.
|
| MCP endpoint or manifest is discoverable mcp_descriptor_discoverable |
NOT VERIFIED |
No documented, single well-known path exists for MCP manifest discovery, and no linked/referenced MCP manifest was found on the homepage or in llms.txt -- this check cannot distinguish absence from an undiscoverable-by-convention manifest, so it is reported as Not verified rather than guessed.- In plain terms
- AI agents can discover your MCP tools
- What it tests
- A documented/well-known MCP endpoint or server manifest is publicly discoverable and its manifest parses as valid JSON with the expected MCP shape. This checks discoverability only — a real MCP handshake and tool-quality review is performed manually during paid fulfilment and recorded separately, never assumed from descriptor presence alone.
- Standard
- Model Context Protocol specification (well-known manifest convention)
- How ARC checks it
- Looks for a published descriptor; presence alone does not prove behaviour
- What ARC attempted
- ARC ran this check as defined (A documented/well-known MCP endpoint or server manifest is publicly discoverable and its manifest parses as valid JSON with the expected MCP shape. This checks discoverability only — a real MCP handshake and tool-quality review is performed manually during paid fulfilment and recorded separately, never assumed from descriptor presence alone.) and recorded: No documented, single well-known path exists for MCP manifest discovery, and no linked/referenced MCP manifest was found on the homepage or in llms.txt -- this check cannot distinguish absence from an undiscoverable-by-convention manifest, so it is reported as Not verified rather than guessed.
- Why it could not be verified
- No reliable discovery convention.
- What would resolve it
- A published, documented location or a direct endpoint reference from you would allow this to be checked.
- Applicability
- Applicable. Applies only when the site exposes a delegated or protected agent-access surface, for example an OAuth-protected API, an MCP or A2A service with protected tools, customer accounts that agents can act within, or another authenticated agent-access surface.
- Why it matters
- Descriptor discoverability is a prerequisite for any agent to even attempt a connection, but presence alone does not prove the server actually works.
- Scoring
- Not included in scoring: not verified checks carry no points. NOT VERIFIED means ARC could not reach a conclusion; it is not a failure.
|
| Agent Skills are discoverable and well-formed agent_skills_discoverable |
NOT VERIFIED |
No publicly documented, standard discovery location exists yet for an Agent Skills manifest on a public website; automating this would require inventing a non-standard convention, so it is always reported as Not verified pending a documented standard.- In plain terms
- Your agent skills are published correctly
- What it tests
- Where an Agent Skills manifest is published at its documented location, checks it parses and includes usable capability descriptions.
- Standard
- Agent Skills convention (emerging, non-standard)
- How ARC checks it
- Looks for a published descriptor; presence alone does not prove behaviour
- What ARC attempted
- ARC ran this check as defined (Where an Agent Skills manifest is published at its documented location, checks it parses and includes usable capability descriptions.) and recorded: No publicly documented, standard discovery location exists yet for an Agent Skills manifest on a public website; automating this would require inventing a non-standard convention, so it is always reported as Not verified pending a documented standard.
- Why it could not be verified
- No reliable discovery convention.
- What would resolve it
- A published, documented location or a direct endpoint reference from you would allow this to be checked.
- Applicability
- Applicable. Applies only when the site exposes a delegated or protected agent-access surface, for example an OAuth-protected API, an MCP or A2A service with protected tools, customer accounts that agents can act within, or another authenticated agent-access surface.
- Why it matters
- Malformed or vague skill descriptions are effectively invisible to an agent trying to decide when to use them.
- Scoring
- Not included in scoring: not verified checks carry no points. NOT VERIFIED means ARC could not reach a conclusion; it is not a failure.
|
| A2A agent card is valid a2a_agent_card_valid |
NOT APPLICABLE |
No A2A agent card was found at /.well-known/agent-card.json.- In plain terms
- Your agent-to-agent identity card is correct
- What it tests
- /.well-known/agent-card.json is fetchable and validates against the expected A2A shape (identity, endpoints, skills, security metadata).
- Standard
- A2A Protocol specification (/.well-known/agent-card.json)
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why ARC treated it as outside scope
- The condition that makes this check applicable was not met for this site. ARC recorded: No A2A agent card was found at /.well-known/agent-card.json.
- What makes it applicable
- Applies only when the site exposes a delegated or protected agent-access surface, for example an OAuth-protected API, an MCP or A2A service with protected tools, customer accounts that agents can act within, or another authenticated agent-access surface.
- Why it matters
- A malformed agent card breaks discovery for any A2A-capable agent trying to interoperate with this service.
- Scoring
- Not included in scoring: not applicable checks carry no points and do not affect the category score.
|
| Public transaction interface is documented public_transaction_interface_documented |
PASS |
A documented public transaction interface was found.- What it tests
- Same OpenAPI/API-docs signal as openapi_documentation_valid, presented as a transaction-readiness signal. Evaluated only where the business intends agent-executed transactions (declared policy) or already publishes an interface; a commerce site with no declared intent is NOT VERIFIED and one that declared agents must not transact is Not applicable. Having a transaction API is never assumed to be the goal.
- Standard
- OpenAPI Specification
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. A documented public transaction interface was found.
- Scoring
- Included in scoring.
|
IDENTITY, AUTHENTICATION & AGENT ACCESS · 2 CHECKS
| CHECK |
RESULT |
EVIDENCE RECORDED BY ARC |
|---|
| OAuth Authorization Server Metadata is discoverable oauth_authorization_server_metadata |
NOT APPLICABLE |
No OAuth Authorization Server Metadata was found.- In plain terms
- Delegated agent access is discoverable
- What it tests
- /.well-known/oauth-authorization-server is fetchable and validates per RFC 8414. Whether it is expected is decided from the business surface and declared agent policy, not from the file's presence: Not applicable when the established surface is public/enquiry-only or the business declared agents may not act; NOT VERIFIED when the surface or intent cannot be established; evaluated when an authenticated/agent-facing surface or declared agent intent exists.
- Standard
- RFC 8414
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why ARC treated it as outside scope
- The condition that makes this check applicable was not met for this site. ARC recorded: No OAuth Authorization Server Metadata was found.
- What makes it applicable
- Applies only when the business surface ARC established includes an authenticated or consequential action surface and the business intends agents to act on it (declared), or an agent-facing interface is published or demonstrably requires authentication. A public or enquiry-only site, or one that declares agents must not act, is outside scope.
- Why it matters
- Standard OAuth metadata lets an agent discover authorization/token endpoints and scopes without out-of-band instructions.
- Scoring
- Not included in scoring: not applicable checks carry no points and do not affect the category score.
|
| OAuth Protected Resource Metadata is discoverable oauth_protected_resource_metadata |
NOT APPLICABLE |
No OAuth Protected Resource Metadata was found.- In plain terms
- Protected agent resources are discoverable
- What it tests
- /.well-known/oauth-protected-resource is fetchable and validates per RFC 9728. Applicability follows the business surface and declared intent (see oauth_authorization_server_metadata). A FAIL requires proof: an agent-facing endpoint that demonstrably answered HTTP 401 while advertising no discovery path.
- Standard
- RFC 9728
- How ARC checks it
- Fetches the content and parses it against the expected structure
- Why ARC treated it as outside scope
- The condition that makes this check applicable was not met for this site. ARC recorded: No OAuth Protected Resource Metadata was found.
- What makes it applicable
- Applies only when the business surface ARC established includes an authenticated or consequential action surface and the business intends agents to act on it (declared), or an agent-facing interface is published or demonstrably requires authentication. A public or enquiry-only site, or one that declares agents must not act, is outside scope.
- Why it matters
- Lets an agent locate the authorization server responsible for a protected resource without prior knowledge.
- Scoring
- Not included in scoring: not applicable checks carry no points and do not affect the category score.
|
CONVERSION & AGENTIC COMMERCE · 3 CHECKS
| CHECK |
RESULT |
EVIDENCE RECORDED BY ARC |
|---|
| Clear conversion action is visible clear_conversion_action_visible |
PASS |
Action language/link detected- What it tests
- The homepage's text or links contain contact/book/buy/order/quote/signup/schedule language.
- Standard
- none — text heuristic
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. Action language/link detected.
- Scoring
- Included in scoring.
|
| Pricing or offer context is visible pricing_offer_context_visible |
NEEDS ATTENTION |
No public pricing/offer language detected (may be an intentional sales-led model, not a defect)- In plain terms
- Pricing or offer information is visible
- What it tests
- The homepage's text contains 'pricing', 'price', or 'offer' language.
- Standard
- none — text heuristic
- How ARC checks it
- Reads the visible page text for the expected signal
- Why this result
- Something relevant is present but incomplete, informational or constrained: No public pricing/offer language detected (may be an intentional sales-led model, not a defect). It earns partial credit and is reported as needing attention rather than as a failure.
- Applicability
- Applicable. Applies to every site, and is scored once the evidence it depends on exists.
- Why it matters
- Helps an agent compare options quickly; a 'contact sales' model is a legitimate choice, not a defect by itself.
- Scoring
- Included in scoring.
|
| Emerging agentic-commerce protocol signals emerging_commerce_protocol_signals |
PASS |
An emerging agentic-commerce protocol signal was found.- What it tests
- Detects publicly documented support for x402, Universal Commerce Protocol, Agentic Commerce Protocol, AP2, or Machine Payments Protocol. Informational only (scoring_weight 0): it moves no score in either direction -- support for a payment protocol is evidence of a channel, not of readiness.
- Standard
- x402 / UCP / ACP / AP2 / Machine Payments Protocol (all emerging, non-standardized)
- Why this result
- The evidence ARC recorded satisfies the condition this check looks for. An emerging agentic-commerce protocol signal was found.
- Scoring
- Included in scoring.
|
Fictitious sample · made-up data
B
Page sample evidence
The pages ARC fetched for this assessment. 9 of 12 were evaluated; 3 could not be reached.
| PAGE TYPE |
URL |
STATUS |
OBSERVATION |
|---|
| homepage |
https://willow-thread.example/ |
EVALUATED |
Primary scanned page; full ARC check set applied here. |
| documentation |
https://willow-thread.example/support/billing/ |
EVALUATED |
Fetched and reviewed for representative-page accessibility. |
| ecommerce_product |
https://willow-thread.example/catalogue/ |
EVALUATED |
Fetched and reviewed for representative-page accessibility. |
| developer_api |
https://willow-thread.example/sitemap-stats_api-1.xml |
EVALUATED |
Reachable, but visible content was too sparse to meaningfully evaluate (may be client-rendered). |
| other_content_page |
https://willow-thread.example/privacy/ |
EVALUATED |
Fetched and reviewed for representative-page accessibility. |
| other_content_page |
https://willow-thread.example/terms/ |
EVALUATED |
Fetched and reviewed for representative-page accessibility. |
| other_content_page |
https://willow-thread.example/discover/ |
EVALUATED |
Fetched and reviewed for representative-page accessibility. |
| other_content_page |
https://willow-thread.example/tags/ |
EVALUATED |
Fetched and reviewed for representative-page accessibility. |
| other_content_page |
https://willow-thread.example/regions/north-america/ |
EVALUATED |
Fetched and reviewed for representative-page accessibility. |
| other_content_page |
https://willow-thread.example/regions/europe/ |
UNREACHABLE |
Not reachable during this audit (status: 429). |
| other_content_page |
https://willow-thread.example/regions/asia/ |
UNREACHABLE |
Not reachable during this audit (status: 429). |
| other_content_page |
https://willow-thread.example/new/ |
UNREACHABLE |
Not reachable during this audit (status: 429). |
Fictitious sample · made-up data
11
Methodology, versions & limitations
ASSESSMENT RECORD
REPORT ID
SAMPLE-ARC-REPORT-FICTITIOUS
SCAN
SAMPLE-NO-LIVE-SCAN
GENERATED
12 October 2026
CURRENT THROUGH
26 October 2026 — ARC results are treated as current for 14 days; after that a rescan is recommended and this document remains a historical record.
SCANNER / REGISTRY
v2.0.0 / v2.0.0
SCORING / SCHEMA
v2.0.0 / 1.0
COVERAGE CLASS
PARTIAL — Partial coverage — one or more pages in the sample could not be reliably assessed; treat findings as directional for those pages.
HOW THE SCORE IS CALCULATED
Each ARC category is weighted equally. Within a category, a check's contribution reflects its registry scoring weight, scaled down for conditional and emerging-informational maturity checks. REACH and CAPABILITY pillar scores are reported separately, never blended beyond the overall score. Severity is a separate axis from points and drives remediation priority. Not verified and Not applicable checks are excluded from scoring entirely.
LIMITATIONS
- Scores reflect the site's configuration as observed at scan time; websites, AI platforms and ARC's own registry change over time.
- Partial coverage — one or more pages in the sample could not be reliably assessed; treat findings as directional for those pages.
- 6 operator-verified checks remain pending and are reported as NOT VERIFIED. Nothing in this report asserts those behaviours work.
- NOT VERIFIED and NOT APPLICABLE results are excluded from scoring; they are neither silent passes nor failures.
- Findings are evidence of observable signals, not a guarantee of any ranking, traffic, revenue or AI-recommendation outcome.
- No agent credentials or secrets were stored during this assessment.
Fictitious sample · made-up data
Next steps
This assessment is a snapshot with an evidence trail. Every finding carries an ARC test ID, so the same tests can be re-run after remediation to verify the change against the criteria used here.
01
Review the roadmap with your technical owner
4 prioritised opportunities, each with its prerequisites and the ARC tests it is verified against.
02
Book a scope review
2 items require a scope review before duration and price can be confirmed. Pricing to be confirmed following scope review.
03
Complete operator verification
6 advanced checks require live operator verification. Results are appended to this report as a new version when recorded.
04
Re-test after changes
A new ARC scan re-runs the same registry (v2.0.0) so movement is measured, not asserted.