The agent successfully assembled a comprehensive explanation of Capian by working around direct website access blocks (Cloudflare 403s) and relying entirely on web search results and cached third-party sources. The agent delivered well-structured answers on what Capian does (expert-driven UX audit tool), who it's for (cross-functional teams), basic pricing ($39/month start), and positioning versus alternatives, but had to acknowledge significant gaps on pricing tier details, support levels, and scale limits that the blocked website would have clarified.

• ›The agent could not fetch capian.co directly (step [1] returned Cloudflare 403), forcing reliance on search results rather than primary sources—75% of fetches came from web search.
• ›Search results (steps [3], [4], [7], [8], [10], [13], [14], [16], [17], [18]) returned URLs to Capian's own pages (pricing, accessibility, integrations, schools, blog) but these pages were themselves also blocked (steps [5], [6], [9]), so the agent could only use search snippets and third-party review aggregators (Capterra, SaasWorthy, DevsQuad, CPOClub).
• ›The agent explicitly acknowledged three categories of missing information: pricing tier feature breakdown, target market emphasis clarity, and support/performance limits—all of which would have been on the blocked Capian site's pricing and help pages.
• ›Capian's site is not agent-ready in practice: it enforces strict Cloudflare access controls that block automated scraping, making the product discoverable only through search aggregators and reviews rather than via direct API or structured metadata (no llms.txt, no accessible schema data mentioned).
• ›The agent cited 10 sources in its final response, but none were direct fetches from capian.co itself; all were either search result snippets or URLs inferred from prior knowledge—demonstrating high reliance on web search inference rather than primary documentation.