The agent assembled a reasonably complete explanation of Astra by combining direct site fetches (homepage, pricing, features pages) with web search results and third-party review comparisons. However, the site's own pages returned truncated HTML that did not expose full content, forcing the agent to rely heavily on search results and prior knowledge to construct the answer. The site itself was not agent-ready; critical information like pricing details and feature comparisons had to be sourced from external reviews rather than the official pages.

• ›Steps [1], [3], and [4] fetched the homepage, pricing, and features pages directly from wpastra.com, but all returned truncated HTML (4400+ chars cut off) that prevented the agent from extracting complete information. The agent explicitly flagged this issue in turn [5].
• ›Steps [6], [7], and [8] used web search to compensate for incomplete site content. Search result snippets and third-party reviews (ManyTools, PluginTheme.net, ThemeIsle, etc.) provided the actual pricing tiers ($59-69/year, $129/year), feature comparisons (template count, multi-builder support), and competitive positioning that the official site pages did not render.
• ›The agent was honest about gaps: it noted that specific Pro vs. free feature distinctions, exact pricing for bundled plans (Mini Agency, Business), and case studies were not found or not visible. This indicates the site either does not surface these details in easily-parsed form, or the truncation prevented discovery.
• ›The site's official pages are client-side rendered or heavily JavaScript-dependent, resulting in truncation when fetched directly. The agent had to rely on 67% prior knowledge (cached/inferred understanding of Astra) and external reviews rather than native site content, which is a poor signal for agent-readiness.