Agent Interactivity
Learn how GEOAudit checks for WebMCP manifests, Action schemas, OpenAPI specs, and other machine-actionable interfaces for AI agent interaction.
What We Check
GEOAudit checks whether your site provides machine-actionable interfaces for AI agents. We look for WebMCP (Model Context Protocol) manifest files, Action schema markup, OpenAPI/Swagger specification availability, API documentation links, deep linking support, webhooks or callback URLs, form action endpoints, and other interfaces that allow AI agents to not just read but interact with your site. This is the cutting edge of AI agent readiness.
How We Score
Agent Interactivity carries a 5% weight in the overall score. Each check produces pass, warn, or fail. Key assessments include: WebMCP manifest detection, Action schema presence, OpenAPI spec availability, API documentation linkage, and deep linking capability. This category has the most forward-looking checks — many sites score low here because these technologies are emerging.
Why It Matters
The future of AI is agentic — AI systems that take actions on behalf of users. WebMCP manifests tell AI agents what actions are available on your site. Action schemas define specific operations. OpenAPI specs let AI agents understand and use your APIs. Sites that support agent interactivity will be first to benefit as AI agents evolve from information retrieval to task execution, enabling booking, purchasing, and data querying through AI.
How to Improve
Create a WebMCP manifest file (.well-known/mcp.json) describing available actions. Add Action schema markup for key interactive elements. If you have an API, publish an OpenAPI specification. Document API endpoints clearly. Support deep linking so AI agents can link to specific content or actions. Consider building webhook endpoints for AI agent callbacks. This category represents emerging standards — early adoption gives you a competitive edge.
Frequently Asked Questions
What is WebMCP (Model Context Protocol)?
WebMCP is an emerging standard that allows websites to describe their capabilities to AI agents in a machine-readable format. A WebMCP manifest tells AI agents what actions they can perform on your site, like searching, booking, or purchasing.
Do I need an API to score well in this category?
No — you can score well with Action schema markup and deep linking even without a full API. However, having an OpenAPI specification significantly boosts your score and future-proofs your site for agentic AI interactions.
Is agent interactivity important right now?
It's becoming important fast. AI agents are evolving from information retrieval to task execution. Sites that support agent interactivity now will be first to benefit as ChatGPT, Claude, and other AI assistants add action capabilities.
What's an Action schema?
Action schema (Schema.org/Action) defines specific actions available on a page, like SearchAction, BuyAction, or OrderAction. These tell AI agents what interactive operations are possible, enabling them to take actions on behalf of users.
Ready to optimize for AI?
Start scanning your pages for free — no account required for the Chrome extension. Or sign up for the full dashboard experience.