The agent successfully gathered a comprehensive overview of Qarnot by combining website fetches with web search results. The main website pages (English homepage and pricing) were reachable but content-thin in HTML, requiring the agent to supplement with search-sourced news articles and external documentation to build a complete picture. The site's sustainability positioning and heat-recovery differentiator were well-documented on a dedicated page; pricing details and specific hardware offerings remained opaque and required sales contact.

• ›Step [5] (qarnot.com/en/pricing) fetched successfully but returned a JavaScript-heavy page that did not render actual pricing tiers or per-unit costs in the HTML—the agent correctly flagged this as 'opaque' and noted users must contact sales.
• ›Step [4] (qarnot.com/en) and step [17] (qarnot.com/en/sustainability) returned content confirming the company's core value props (HPC for CAE, heat recovery, European sovereignty), but the agent had to infer details about competitor positioning and software support from external search results rather than from the website itself.
• ›Step [13] search results (CES 2025 announcements, Scaleway acquisition news) provided critical context about the heat-recovery technology's impact (80% carbon reduction, 50% cost savings) and the July 2026 acquisition—information that shaped the final explanation but was sourced from third-party outlets, not from qarnot.com.
• ›The website's main pages exist and are navigable, but lack granular product specs, transparent pricing, and detailed competitor comparisons; the agent had to assemble the answer from fragmented on-site content plus external sources, reducing the site's machine-readability for this task.