Website diagnostic for the agentic web
Can AI agents actually use your website?
See whether AI agents can understand your business and identify actions such as bookings, quote requests, forms and enquiries.
We scan publicly accessible pages to assess how ready your website is for AI agents. Free scan. No signup required. Scan up to five public pages. No account required. Read the methodology, the live benchmark research or the verified directory.
- Pages read
- Up to 5
- Checks run
- 32 rules
- Scoring
- Rule-based
- Account needed
- None
Verification
Websites with WebMCP tools observed live
Each of these profiles is anchored to the exact page we tested in a compatible browser. Source-level signals are listed separately from tools observed live.
Context
AI is changing how customers use websites
Websites have traditionally been designed around one thing: a person navigating with their eyes and a mouse. Layout, imagery and visual hierarchy carry most of the meaning, and that works well enough when a human is doing the reading.
Increasingly, software reads your site too. Search crawlers, comparison tools and AI assistants extract your business details, work out what you offer, and pass that on to the person asking. They rely on structured information and clearly defined actions rather than visual cues.
People still browse websites the normal way, and AI agents have not replaced that. But when software can't tell what you sell, where you work or how to contact you, you're absent from an increasing share of those conversations. The fixes are the same ones that make the site clearer for people.
Method
What we check
Agent discovery
Can an AI agent identify the relevant capabilities on your site?
AG AgentReady checks the public signals that help software discover and interpret the actions your website makes available.
- Public pages and agent-facing discovery signals
- Structured and semantic information that supports interpretation
- WebMCP source-level references where present
Agent actions
What useful actions can an AI agent identify?
We compare the actions available to people with the structured capabilities available to AI agents, so missing or unclear agent actions can be identified.
- Quote, booking, enquiry and contact actions
- Human-facing actions compared with agent-facing tools
- Evidence-backed gaps where no equivalent tool is observed
Runtime verification
What capabilities are actually registered on the tested page?
Where compatible runtime validation is available, AG AgentReady loads the exact page and inspects the WebMCP tools it registers.
- Exact page tested
- Registered tool names, descriptions and available input schemas
- Registered and inspected — not executed
Know what your website should let AI agents do
Create a tailored WebMCP tool plan based on your website, platform, integrations and customer actions.
Example report
See what capabilities your webpage exposes to AI agents
AG AgentReady checks a specific webpage for live WebMCP tools, inspects the available tool definitions and compares them with the actions a customer can take.
Illustrative example
https://example.com/book
Exact page tested
https://example.com/book
Runtime validation
Completed
Qualifying WebMCP tools observed
2
Inspection status
Registered and inspected — not executed
Example observed-tool inventory
Tool name
search_services
Description
Finds services matching a customer’s requirements.
Available inputs
service_typelocationpreferred_date
Tool name
request_quote
Description
Creates a structured request for a service quotation.
Available inputs
service_typeproject_detailscontact_email
What the report examines
- WebMCP detection state
- Exact page and runtime-validation scope
- Tool names, descriptions and available input schemas
- Agent-action coverage
- Missing or incomplete agent-facing capabilities
- Evidence, limitations and implementation guidance
AG AgentReady reads the WebMCP registry on the exact page tested. It inspects available tool descriptions and input schemas, but never calls a tool, submits information or completes an action.
Illustrative example. Exact-page results are anchored to the page tested. Where available, the report also shows WebMCP capabilities observed on other selected pages.
Pricing
Pay once, no subscription
Free Agent Readiness Snapshot
$0
No signup required
- WebMCP detection state for the page we test
- The exact page tested and the validation scope
- Observed tool count and tool names
- Headline Agent Readiness verdict
- Full observed-tool inventory with input schemas
- Evidence, limitations and prerequisites
- Proposed tool specifications and implementation steps
Complete Agent Readiness Report
$9
One-time payment
Available after your free scan
- WebMCP detection state, exact tested page and validation scope
- Complete observed-tool inventory with descriptions and input schemas
- Registered, inspected and not-executed status for every tool
- Evidence and limitations behind each observation
- Machine-readable prerequisites
- Agent-action coverage analysis and evidence-backed capability gaps
- Evidence-supported proposed tool specifications
- Suggested tool names, descriptions, inputs and response shapes
- Authentication, consent, privacy and safety considerations
- Progressive-enhancement guidance and implementation steps where evidence supports them
- Runtime validation steps
- Complete PDF download
One $9 payment unlocks the Complete Agent Readiness Report: every observation with its evidence, and the proposed tool specifications.
All paid products are automated digital reports and action plans delivered online immediately after purchase. No consulting, implementation, or done-for-you services are included.
FAQ
Questions
Run your free website snapshot
One URL, no account, up to five public pages.