AX 감사 보고서

https://karumi.ai

August 12, 20261405ms response

58AX score / 100

Needs work

Readiness diagnosis

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

0개 심각

5 warnings

20개 통과

Current constraintAgent Integration13%

Five readiness layers

Readiness orbit

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

Evidence profile

25 checks recorded

Passed vs. material gaps

Critical

Warnings

Passed

Discovery

4 / 13 weighted points

31%

Identity

16.8 / 22 weighted points

76%

Access

28.3 / 31 weighted points

91%

Agent Integration

Current constraint

3.1 / 24 weighted points

13%

User Experience

5.3 / 10 weighted points

53%

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

에이전트 상호작용

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>, <footer>, multiple <h1> tags (use only one)

콘텐츠 품질

Review

Only 0% of images have alt text (0/118)

콘텐츠 품질

Review

Explore the audit

Findings and fixes

에이전트 상호작용

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 1 JSON-LD structured data block

Business schema found: Organization

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.

sitemap.xml found and valid

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

Only 0% of images have alt text (0/118)
Add descriptive alt attributes to all <img> tags. Example: <img alt="Description of image content" src="..." />

Sufficient text content available for AI extraction

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.

Contact information detected (phone, address)

Fast response time (1405ms)

Large HTML payload (1051KB)

Response is compressed (br)

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

Entry URL redirects before serving — adds latency for agents

No author/publisher in schema — weak authorship signal
Add author and publisher to your Schema.org JSON-LD to establish credibility:
"publisher": { "@type": "Organization", "name": "Your Brand" }

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"

No current year found — the site may look stale to agents

HTTPS enabled

SSL certificate valid

Privacy policy link found

Terms of service link found

karumi.ai

https://karumi.ai

Readiness orbit

Top Priorities

Findings and fixes

All Findings

Turn audit findings into an agent-ready funnel.

Related AX Scores