Agent-Ready Audit

17 Detection Functions

Score your site across four categories. See how easily AI agents can navigate and complete tasks.

Score Categories

Each category has a weighted percentage that feeds into your overall AI-friendliness score.

Findability

Can agents locate elements by intent, not just selector?

35%

weight

ARIA labels presentDescriptive button textMeaningful link textForm labels associated

Stability

Will selectors survive DOM changes and rebuilds?

30%

weight

Stable IDsData attributesNo dynamic class namesConsistent structure

Accessibility

ARIA labels, roles, and keyboard navigation

20%

weight

ARIA rolesFocus managementKeyboard navigableScreen reader compatible

Semantics

Structured data, heading hierarchy, and metadata

15%

weight

JSON-LD presentOpenGraph tagsProper heading hierarchyllms.txt available

Grade Scale

Your overall score maps to a letter grade from A (excellent) to F (failing).

90-100

Excellent - AI agents navigate effortlessly

80-89

Good - Minor issues may slow agents

70-79

Fair - Some barriers to agent navigation

60-69

Poor - Significant agent navigation issues

<60

Failing - Major barriers to AI agents

Detection Functions

All 17 checks in an Agent-Ready Audit. Organized by category and severity.

Function	Category	Severity	Description
`unlabeledButtons`	Findability	high	Buttons without accessible labels or text content
`unlabeledInputs`	Findability	high	Form inputs without associated labels
`genericButtonText`	Findability	medium	Buttons with generic text like "Submit", "Click", "Go"
`missingLinkText`	Findability	high	Links without descriptive text (icon-only, empty)
`hiddenSelectInputs`	Findability	high	Native selects hidden by custom dropdown replacements
`customDropdownsNoAria`	Accessibility	high	Custom dropdowns without ARIA combobox pattern
`missingFormLabels`	Findability	high	Form fields without programmatic label association
`dynamicClassNames`	Stability	medium	CSS classes that appear generated or hashed
`noStableIds`	Stability	medium	Interactive elements without stable ID attributes
`missingJsonLd`	Semantics	low	No JSON-LD structured data on the page
`missingOpenGraph`	Semantics	low	Missing or incomplete OpenGraph metadata
`missingLlmsTxt`	Semantics	medium	No llms.txt file at site root
`poorHeadingHierarchy`	Semantics	medium	Heading levels skip or are out of order
`infiniteScrollPattern`	Stability	high	Infinite scroll without pagination fallback
`noSkipLinks`	Accessibility	low	Missing skip-to-content links
`focusTrapIssues`	Accessibility	high	Modal/dialog focus trap problems
`missingAltText`	Findability	medium	Images without alt text attributes

CLI Reference

Run an Agent-Ready Audit from the command line. Several options are available.

# Basic audit

npx cbrowser agent-ready-audit https://example.com

# Audit with patches (remediation suggestions)

npx cbrowser agent-ready-audit https://example.com --patches

# Audit specific persona strategy

npx cbrowser agent-ready-audit https://example.com --persona task-completion-agent

# JSON output for CI/CD integration

npx cbrowser agent-ready-audit https://example.com --format json

# MCP tool invocation

{
  "tool": "agent_ready_audit",
  "params": {
    "url": "https://example.com",
    "device": "mobile"
  }
}

# Mobile AI-readiness audit

npx cbrowser agent-ready-audit https://example.com --device mobile

Sample Report

A sample Agent-Ready Audit report for a typical e-commerce site.

example-store.com

Audited 2 minutes ago

Score: 84/100

Findability

91/100

Weight: 35%

Stability

78/100

Weight: 30%

Accessibility

85/100

Weight: 20%

Semantics

72/100

Weight: 15%

Issues Found (8)

3 buttons with generic labels ("Submit", "Click Here")

2 custom dropdowns without ARIA combobox pattern

Missing llms.txt at site root

Dynamic class names on 12 elements

No skip-to-content link

Found Issues? Get Patches

Use the Remediation Engine to get code patches. Each patch includes before/after comparisons and effort/impact scoring.

View Remediation Engine Generate llms.txt