Full Answer

Lighthouse 13.3 introduced the Agentic Browsing audit category alongside its existing performance, accessibility, and SEO checks. The audit evaluates whether a website exposes the structured interfaces that AI agents need to navigate, query, and act on a site without human-style browsing.

The five dimensions scored are: structured data coverage (do product and content pages carry complete schema markup), API surface discoverability (does the site expose endpoints or feeds that agents can consume programmatically), navigation predictability (can an agent traverse the site's information architecture without JavaScript rendering), authentication surface (are there machine-readable paths for agent identity and permissioning), and action completability (can an agent initiate a transaction or form submission through structured inputs).

For WordPress and WooCommerce sites, several of these dimensions require deliberate configuration. WooCommerce generates basic JSON-LD product schema by default, but it omits attributes like shipping estimates, return policies, and review aggregates that agents weight heavily. The REST API provides programmatic access, but most stores leave it unauthenticated and undocumented from an agent's perspective. Adding an llms.txt file, completing product schema fields, and exposing a structured sitemap that agents can parse gives a WordPress site a meaningful head start.

Passing the audit is not mandatory today — but the signals it checks are already used by AI shopping agents to rank which stores they engage with. Treating it as a readiness checklist rather than a compliance gate is the pragmatic approach for stores that want to stay visible as agent-driven commerce grows.