Rapport d'audit AX

https://cosine.sh

August 12, 2026687ms response

59AX score / 100

Needs work

Readiness diagnosis

This site needs work before agents can use it reliably. Address the priority findings to improve readiness.

0 critique

6 warnings

22 réussi

Current constraintAgent Integration7%

Five readiness layers

Readiness orbit

The shape reveals where an agent journey breaks down, from finding the site to completing useful work.

Evidence profile

28 checks recorded

Passed vs. material gaps

Critical

Warnings

Passed

Discovery

3.8 / 13 weighted points

29%

Identity

15.7 / 22 weighted points

71%

Access

31 / 31 weighted points

100%

Agent Integration

Current constraint

1.7 / 24 weighted points

User Experience

6.7 / 10 weighted points

67%

Start here

Top Priorities

Select the highest-impact Findings to include in your Fix Plan.

No llms.txt file found — AI agents cannot discover your services

Interaction avec l'agent

Review

No sitemap.xml found

Interaction avec l'agent

Review

No Schema.org potentialAction — agents can't tell what they can do on your site

Actionability

Review

No form or direct contact path — agents have no way to convert here

Actionability

Review

Missing semantic HTML: <main>, multiple <h1> tags (use only one)

Qualité du contenu

Review

Explore the audit

Findings and fixes

Interaction avec l'agent

Actionability

All Findings

robots.txt allows all major AI crawlers

Page returns HTTP 200

Page content available without JavaScript

No CAPTCHA or bot protection detected

Found 3 JSON-LD structured data blocks

Business schema found: Organization

Schema has 4 fields — consider adding more for better AI comprehension

OpenGraph tags complete (title, description, image)

Meta tags complete (title, description, canonical)

No llms.txt file found — AI agents cannot discover your services
Create a /llms.txt file at your site root:

# Your Business Name

> Brief description of your business.

## Services
- [Service 1](/service-1): Description
- [Service 2](/service-2): Description

See https://llmstxt.org for the full spec.

No MCP server or AI agent endpoint detected — agents cannot interact with your business
Set up a Model Context Protocol (MCP) server so AI agents can interact with your business directly. This allows agents to book appointments, check availability, and more.

With Dashform, you can add MCP support to your website in minutes — no coding required.

Alternatively, create /.well-known/ai-plugin.json for ChatGPT plugin compatibility.

No sitemap.xml found
Create a /sitemap.xml file listing all your important pages. Also declare it in robots.txt:
Sitemap: https://yoursite.com/sitemap.xml

No Schema.org potentialAction — agents can't tell what they can do on your site
Add a potentialAction to your Schema.org JSON-LD so agents can act, e.g.:
"potentialAction": {
  "@type": "ReserveAction",
  "target": "https://yoursite.com/book"
}

No scheduling link found — agents can't book time with you directly

No form or direct contact path — agents have no way to convert here
Add an agent-usable conversion path — a contact form, a booking widget, or a Dashform AI funnel agents can complete on the visitor's behalf.

CTA buttons have descriptive labels

Missing semantic HTML: <main>, multiple <h1> tags (use only one)
Add missing semantic HTML elements: <main>, multiple <h1> tags (use only one)

98% of images have alt text (87/89)

Sufficient text content available for AI extraction

Heading levels skip (e.g., h1 directly to h3)

2 social media links found

Schema.org sameAs links found (cross-platform identity)

No Google Maps or Business Profile presence detected — AI agents may not verify your location
Claim your Google Business Profile at https://business.google.com and embed a Google Maps link or iframe on your page. AI agents use Google Maps data to verify business locations.

Limited contact information — add phone, email, and address

Fast response time (687ms)

Lean HTML payload (239KB)

Response is compressed (br)

Caching headers present (Cache-Control/ETag/Last-Modified)

Author/publisher identity declared in schema

No review/rating schema — agents can't surface your reputation

Partial security headers (HSTS only)

No dateModified/datePublished in schema — agents can't judge freshness
Add dateModified/datePublished to your Schema.org JSON-LD so agents can judge freshness:
"dateModified": "2026-01-15"

Footer/copyright year is current

HTTPS enabled

SSL certificate valid

Privacy policy link found

Terms of service link found

cosine.sh

https://cosine.sh

Readiness orbit

Top Priorities

Findings and fixes

All Findings

Turn audit findings into an agent-ready funnel.

Related AX Scores