AI agent readiness checker
The next wave after AI search is AI agents: assistants that do not just read your site but use it. This checker tracks how ready you are, before readiness becomes table stakes.
Why it matters for AI search
WebMCP is the W3C effort that lets a page declare tools an agent can call: search this catalog, book this slot, submit this form. NLWeb and MCP servers point the same direction. None of it is production yet, which is exactly why GetSlurp tracks it at zero weight instead of pretending it affects your score today.
Some of it already matters. ChatGPT-User and Claude-User visit pages on behalf of users right now, and they interact with whatever your markup exposes. Labeled forms, named fields and clean structure decide whether those visits succeed. We report your readiness today and will weight this category the moment engines ship support.
What GetSlurp checks
This list comes straight from the audit engine, so it always matches what the extension and MCP server actually run.
agent.webmcp-signalsInformationalWebMCP is the W3C effort that lets a page declare tools AI agents can call directly. Early adoption positions the site for agent traffic before it becomes table stakes.
Fix: Nothing required today. Watch the WebMCP spec and add tool declarations once your platform supports them.
agent.mcp-discoveryInformationalSites are starting to expose MCP servers so AI assistants can interact with them directly instead of scraping.
Fix: If your product has an API, consider publishing an MCP server and linking it from your site.
agent.forms-operableInformationalBrowsing agents (ChatGPT-User, Claude-User) already fill and submit forms for users. Labeled, named fields are what make that work.
Fix: Give every form field a label and a name attribute so agents can identify what each field expects.
Try GetSlurp free
Audit any page with 60+ checks across 12 categories. No sign-up, runs on your machine.