Speed Insights
Grade A
99% performance
https://agentcache.run
From Agentcache.run
LCP
428ms
Largest paint
TBT
0ms
Blocking time
CLS
0.000
Layout shift
70
out of 100
93
out of 100
140/150 rule points
86
out of 100
62/72 rule points
85
out of 100
68/80 rule points
69
out of 100
76/110 rule points
Category point breakdown
Biggest score-losing checks
For your AI agent
Share this audit’s URL with ChatGPT, Claude, or your coding agent to turn findings into fixes. Your agent can open the saved evidence below when it needs more detail; these files never start a new scan.
Agent-readable files
Add common public security headers
Why this needs attention
Several browser protection headers are missing or present only as weak, narrow policies.
These headers reduce avoidable browser-side risk and show a baseline of care before launch.
Recommended correction
Evidence
Detected 3 of 5 common public security controls. Present: x-content-type-options, frame protection (x-frame-options), referrer-policy. Missing: strict-transport-security, content-security-policy. Needs strengthening: none.
Priority
Priority 5: fix before sharing the page publicly.
Fix Prompt
Make the primary CTA obvious
Why this needs attention
The page does not have a detectable call to action.
Early visitors need a clear next step, such as trying the product, joining a waitlist, or booking a demo.
Recommended correction
Evidence
Rendered CTA evidence: 0 desktop above-fold action(s) across 0 unique purpose(s); 0 mobile above-fold action(s) across 0 unique purpose(s). 0 repeated occurrence(s) were deduplicated and earned no extra credit.
Priority
Priority 4: fix before sharing the page publicly.
Fix Prompt
Strengthen answer-engine content quality
Why this needs attention
The page is technically discoverable, but its entity clarity, answer-ready copy, structured data, evidence, authorship, or freshness is not yet complete.
Search and answer systems need clear, extractable, verifiable facts in addition to crawl access and metadata.
Recommended correction
Evidence
Role-aware AEO content components: entity clarity 80/100, answer-ready text 70/100, semantic structured data 85/100. Detected role: home (high confidence).
Priority
Priority 6: fix during launch polish.
Fix Prompt
Fix rendered axe accessibility violations
Why this needs attention
The rendered page has accessibility rule violations detected by axe-core.
axe checks the actual browser-rendered page, so these issues can affect people using keyboards, screen readers, or other assistive technology.
Recommended correction
Evidence
axe found 3 violation rule(s); 2 non-contrast rule(s) are scored here, including 0 serious or critical rule(s). Color contrast is scored once by the dedicated contrast check. 1 result(s) require review. Top rules: color-contrast (23), landmark-no-duplicate-contentinfo (1), landmark-unique (1).
Priority
Priority 5: fix before sharing the page publicly.
Affected elements
<a href="/" class="mono" style="font-size: 12.5px; font-weight: 600; text-decoration: none; padding: 4px 14px; border-radius: 999px; background: var(--primary); color: var(--primar
✓ 428 pages found
✓ 428 pages fetched
Document should not have more than one contentinfo landmark
Fix any of the following: Document has more than one contentinfo landmark
Fix Prompt
Review introductory clarity signals
Why this needs attention
One or more heading, introductory-copy, or next-action signals fell outside this heuristic's expected range; this does not by itself prove the page is vague.
The introduction should help visitors understand the purpose of this particular page.
Recommended correction
Evidence
Detected 1 H1 heading(s), 26 H1 characters, 181 introductory supporting-copy characters after excluding bylines and short labels, and 0 actionable CTA(s). The H1 text is treated as a subject label, so an appropriate short place, product, or page name does not lose points. Detected role: home (high confidence).
Priority
Priority 6: fix during launch polish.
Fix Prompt
SEO / AEO
Severity mix: 1 critical, 8 high, 3 medium, 1 low. Coverage: 0 review and 0 not applicable.
Security
Severity mix: 2 critical, 3 high, 0 medium, 0 low. Coverage: 0 review and 1 not applicable.
Accessibility
Severity mix: 0 critical, 3 high, 6 medium, 0 low. Coverage: 0 review and 2 not applicable.
Design
Severity mix: 0 critical, 3 high, 7 medium, 1 low. Coverage: 0 review and 0 not applicable.
SEO / AEO
Severity mix: 1 critical, 8 high, 3 medium, 1 low. Coverage: 0 review and 0 not applicable.
Accessibility
Severity mix: 0 critical, 3 high, 6 medium, 0 low. Coverage: 0 review and 2 not applicable.
Security
Severity mix: 2 critical, 3 high, 0 medium, 0 low. Coverage: 0 review and 1 not applicable.
Design
Severity mix: 0 critical, 3 high, 7 medium, 1 low. Coverage: 0 review and 0 not applicable.
Failed checks
These checks need attention.
Answer-engine content quality
Why this needs attention
The page is technically discoverable, but its entity clarity, answer-ready copy, structured data, evidence, authorship, or freshness is not yet complete.
Search and answer systems need clear, extractable, verifiable facts in addition to crawl access and metadata.
Recommended correction
Evidence
Role-aware AEO content components: entity clarity 80/100, answer-ready text 70/100, semantic structured data 85/100. Detected role: home (high confidence). Failed: earned 31 of 40 points from partial coverage.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 6: fix during launch polish.
Fix Prompt
Schema quality and validity
Why this needs attention
The JSON-LD may parse, but its required facts, page-role relevance, visible-content agreement, or graph connections are incomplete.
Invalid structured data can be ignored by search and answer systems, even when visible page content is strong.
Recommended correction
Evidence
2 of 2 JSON-LD block(s) parsed. Semantic quality is 85/100: completeness 100/100, graph connectivity 0/100, primary types Organization, SoftwareApplication, currently useful search-feature types Organization, SoftwareApplication. Nested types do not count as separate schema coverage. Failed: earned 7 of 8 points from partial coverage.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 10: fix during launch polish.
Fix Prompt
Common security headers
Why this needs attention
Several browser protection headers are missing or present only as weak, narrow policies.
These headers reduce avoidable browser-side risk and show a baseline of care before launch.
Recommended correction
Evidence
Detected 3 of 5 common public security controls. Present: x-content-type-options, frame protection (x-frame-options), referrer-policy. Missing: strict-transport-security, content-security-policy. Needs strengthening: none. Failed: earned 8 of 18 points from partial coverage.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 5: fix before sharing the page publicly.
Fix Prompt
Basic contrast
Why this needs attention
The scanner found text/background color pairs that may be hard to read, or could not verify contrast.
Low contrast makes a launch page feel less polished and can exclude users with low vision.
Recommended correction
Evidence
42 of 233 rendered text color sample(s) across captured viewports missed WCAG contrast targets. Worst rendered ratio: 3.83. Mobile 390px viewport: 22 of 133 failed; worst ratio 3.83:1. Failed: earned 0 of 8 points.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 20: lower-risk cleanup after urgent launch blockers.
Affected elements
add
openai-docs/
openai-docs/
README.md
Fix Prompt
Rendered axe accessibility violations
Why this needs attention
The rendered page has accessibility rule violations detected by axe-core.
axe checks the actual browser-rendered page, so these issues can affect people using keyboards, screen readers, or other assistive technology.
Recommended correction
Evidence
axe found 3 violation rule(s); 2 non-contrast rule(s) are scored here, including 0 serious or critical rule(s). Color contrast is scored once by the dedicated contrast check. 1 result(s) require review. Top rules: color-contrast (23), landmark-no-duplicate-contentinfo (1), landmark-unique (1). Failed: earned 14 of 18 points from partial coverage.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 5: fix before sharing the page publicly.
Affected elements
<a href="/" class="mono" style="font-size: 12.5px; font-weight: 600; text-decoration: none; padding: 4px 14px; border-radius: 999px; background: var(--primary); color: var(--primar
✓ 428 pages found
✓ 428 pages fetched
Document should not have more than one contentinfo landmark
Fix any of the following: Document has more than one contentinfo landmark
Fix Prompt
Visible primary CTA
Why this needs attention
The page does not have a detectable call to action.
Early visitors need a clear next step, such as trying the product, joining a waitlist, or booking a demo.
Recommended correction
Evidence
Rendered CTA evidence: 0 desktop above-fold action(s) across 0 unique purpose(s); 0 mobile above-fold action(s) across 0 unique purpose(s). 0 repeated occurrence(s) were deduplicated and earned no extra credit. Failed: earned 0 of 14 points.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 4: fix before sharing the page publicly.
Fix Prompt
Hero clarity signals
Why this needs attention
One or more heading, introductory-copy, or next-action signals fell outside this heuristic's expected range; this does not by itself prove the page is vague.
The introduction should help visitors understand the purpose of this particular page.
Recommended correction
Evidence
Detected 1 H1 heading(s), 26 H1 characters, 181 introductory supporting-copy characters after excluding bylines and short labels, and 0 actionable CTA(s). The H1 text is treated as a subject label, so an appropriate short place, product, or page name does not lose points. Detected role: home (high confidence). Failed: earned 11 of 13 points from partial coverage.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 6: fix during launch polish.
Fix Prompt
Credibility and proof
Why this needs attention
The page does not expose enough verifiable credibility for its detected role.
Trust signals reduce hesitation when strangers see the product for the first time.
Recommended correction
Evidence
Detected 0 commercial proof signal(s) and 0 source or methodology signal(s) for this home page. Failed: earned 0 of 10 points.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 16: lower-risk cleanup after urgent launch blockers.
Fix Prompt
Rendered composition and spacing
Why this needs attention
Rendered geometry found cramped text, overly long lines, undersized body copy, or too many competing actions in the first viewport.
Consistent spacing helps the page feel deliberate and easier to scan.
Recommended correction
Evidence
Desktop 1366px: 0 unique above-fold action(s), 3 body sample(s) below 14px, 0 cramped line-height sample(s), and 0 overlong line sample(s). Mobile 390px: 0 unique above-fold action(s), 3 body sample(s) below 14px, 0 cramped line-height sample(s), and 0 overlong line sample(s). Repeated CTA labels are deduplicated and do not earn extra credit. Failed: earned 2 of 7 points from partial coverage.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 25: lower-risk cleanup after urgent launch blockers.
Fix Prompt
Visual hierarchy
Why this needs attention
Rendered typography, geometry, and contrast show that the main heading is not distinct or legible enough, or that the hierarchy needs manual review.
Hierarchy tells visitors what to read first and what action to take next.
Recommended correction
Evidence
The rendered H1 is 52px/500 weight versus a 15px/400 median body style (3.47× size). Captured 8 supporting heading(s) and 1 action(s). The main heading has a clear rendered emphasis over body copy. 42 of 233 rendered text samples missed contrast targets, so hierarchy is not considered fully legible. Failed: earned 6 of 9 points from partial coverage.
How FreeScan evaluates this check (opens in a new tab)Priority
Priority 12: fix during launch polish.
Fix Prompt
Page-level AEO readiness
Shared eligibility and content signals that affect every search and answer agent.
Index and citation eligibility
No general noindex directive was detected.
Snippet and answer controls
No general nosnippet or max-snippet:0 control was detected.
Canonical alignment
The canonical resolves to the scanned public URL.
Sitemap coverage and freshness
The exact URL appears in the fetched sitemap with lastmod 2026-09-11.
Structured data
Semantic schema score 85/100. Detected 2 JSON-LD block(s), 0 parse error(s), 5 graph node(s), and primary types Organization, SoftwareApplication. Completeness: 100/100. Graph connectivity: 0/100. A primary schema type matches the detected page role. The schema entity is supported by visible page content.
Public or paywalled content
No isAccessibleForFree structured-data value was detected.
Entity clarity
Entity alignment scored 80/100 for “Give your agents the docs.”. Title/H1 token overlap is 33%; description match detected; identity schema detected.
Answer-ready visible text
Answerability scored 70/100 from 754 main-content words, 5 concise answer block(s), 0 question heading(s), 2 definition(s), 0 list(s), and 0 table(s).
Authorship and accountability
Authorship is not a universal requirement for the detected home page role.
Published or updated date
Freshness metadata is contextual for the detected home page role.
Supporting-source links
Supporting sources are contextual for the detected home page role and do not affect readiness.
Rendered content availability
Rendered audit completed with 113 visible text element(s).
Synthetic search-agent reachability
8 of 8 completed search/discovery User-Agent probes returned reachable content.
Optional LLM-readable summary
https://agentcache.run/llms.txt returned HTTP 200; this is supplemental and does not determine AI-search eligibility.
OpenAIChatGPT searchOAI-SearchBotPASS · ALLOWED · Toggle crawler details
AI search
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
OpenAIOpenAI model trainingGPTBotOPTED OUT · Toggle crawler details
Model training
Robots: blocked by gptbot disallow: / · line 55
Synthetic probe: not applicable to this policy control
OpenAIChatGPT user fetchesChatGPT-UserUNKNOWN · Toggle crawler details
User-requested fetch
Robots: not applicable to user-requested fetches
Synthetic probe: not tested · access is unverified
AnthropicClaude model trainingClaudeBotOPTED OUT · Toggle crawler details
Model training
Robots: blocked by claudebot disallow: / · line 46
Synthetic probe: not applicable to this policy control
AnthropicClaude searchClaude-SearchBotPASS · ALLOWED · Toggle crawler details
AI search
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
AnthropicClaude user fetchesClaude-UserPASS · ALLOWED · Toggle crawler details
User-requested fetch
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
GoogleGoogle Search and AI featuresGooglebotPASS · ALLOWED · Toggle crawler details
Search and discovery
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
GoogleGemini data use and groundingGoogle-ExtendedOPTED OUT · Toggle crawler details
Data-use policy control
Robots: blocked by google-extended disallow: / · line 52
Synthetic probe: not applicable to this policy control
MicrosoftBing and CopilotBingbotPASS · ALLOWED · Toggle crawler details
Search and discovery
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
PerplexityPerplexity searchPerplexityBotPASS · ALLOWED · Toggle crawler details
AI search
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
PerplexityPerplexity user fetchesPerplexity-UserUNKNOWN · Toggle crawler details
User-requested fetch
Robots: not applicable to user-requested fetches
Synthetic probe: not tested · access is unverified
AppleSiri, Spotlight, and Safari searchApplebotPASS · ALLOWED · Toggle crawler details
Search and discovery
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
AppleApple foundation-model trainingApplebot-ExtendedOPTED OUT · Toggle crawler details
Data-use policy control
Robots: blocked by applebot-extended disallow: / · line 37
Synthetic probe: not applicable to this policy control
Mistral AIMistral searchMistralAI-IndexPASS · ALLOWED · Toggle crawler details
AI search
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
Mistral AIMistral user fetchesMistralAI-UserPASS · ALLOWED · Toggle crawler details
User-requested fetch
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
MetaMeta AI model and product datameta-externalagentOPTED OUT · Toggle crawler details
Model training
Robots: blocked by meta-externalagent disallow: / · line 58
Synthetic probe: not applicable to this policy control
MetaMeta AI user fetchesmeta-externalfetcherPASS · ALLOWED · Toggle crawler details
User-requested fetch
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
AmazonAmazon model trainingAmazonbotOPTED OUT · Toggle crawler details
Model training
Robots: blocked by amazonbot disallow: / · line 34
Synthetic probe: not applicable to this policy control
AmazonAlexa and Rufus searchAmzn-SearchBotPASS · ALLOWED · Toggle crawler details
AI search
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
AmazonAlexa user fetchesAmzn-UserPASS · ALLOWED · Toggle crawler details
User-requested fetch
Robots: allowed by * allow: / · line 31
Synthetic probe: reachable · HTTP 200 · 100% sampled-content parity
Common CrawlOpen web datasetsCCBotOPTED OUT · Toggle crawler details
Model training
Robots: blocked by ccbot disallow: / · line 43
Synthetic probe: not applicable to this policy control
Ready combines robots policy, a bounded synthetic User-Agent fetch, indexability, and snippet eligibility. Synthetic probes do not prove vendor-origin traffic. Preference states are training or data-use choices and do not lower the AEO score.
The image refreshes from FreeScan after a completed rescan. Clicking it always opens the newest public report saved for this exact scanned URL.
Scan evidence is included as untrusted data. The implementation brief instructs agents to verify findings before editing and to avoid destructive or speculative changes.
