How agent-ready is WordPress.com?

WordPress.com scored 41 out of 100 on agentability across its homepage, pricing, and documentation pages, indicating that autonomous AI agents will encounter significant friction when attempting to parse, navigate, and operate the site. The platform performs well on machine readability (86/100), meaning its HTML structure is sound, but falls short on principles that help agents understand system behavior and user intent.

The most critical gaps are in transparency (0/100), shadow UI avoidance (15/100), and status communication (35/100). WordPress.com provides no machine-readable explanations of how results are generated, relies heavily on client-side rendering that obscures content from agents, and fails to announce dynamic status updates in accessible formats. Chunking (39/100) and control (47/100) also need attention—content is not structured to answer user questions directly, and long-running operations lack clear cancellation controls.

These scores suggest that while agents can technically read WordPress.com's markup, they struggle to operate the platform confidently, understand system responses, or guide users through multi-step workflows. Addressing the transparency and status gaps would yield the largest improvements in agent-readiness.

WordPress.com can improve agentability by implementing the following changes:

• Wrap all dynamic status messages and notifications in ARIA live regions (aria-live="polite" or role="status") so agents can detect and announce state changes during imports, publishing, or other asynchronous operations.
• Add a visible, clearly labeled pause or cancel button for long-running tasks such as site imports or media uploads, with accessible names like "Cancel import" rather than icon-only controls.
• Restructure documentation and help content so each section begins with a direct answer in the first one to two sentences, and frame headings as questions (for example, "How do I cancel my subscription?" instead of "Cancellation").
• Expose an activity log or audit trail that records system actions in both human-readable and machine-readable formats, such as a visible event feed paired with a JSON endpoint, to help agents understand what the platform has done and why.
• Attach source references, input citations, or confidence indicators to generated or suggested content so agents can assess reliability and trace recommendations back to their origin.