The agent successfully assembled a comprehensive explanation of Vulpo by combining fetches from the homepage, pricing page, FAQ, support page, and the App Store listing. The vulpo.app website itself was minimally navigable—many logical routes (features, about, why) returned 404s—forcing the agent to rely heavily on prior knowledge and the App Store page to fill gaps in pricing details, feature descriptions, and positioning against alternatives.

• ›Steps [1], [2], [4], [10] fetched from vulpo.app but the HTML responses were truncated and the agent could not extract the full rendered content—the site returns raw Next.js markup without visible text. The agent had to infer details from metadata, headers, and prior knowledge rather than parsing page body content.
• ›Step [9] (App Store listing) was the only external source that returned substantive human-readable content describing features (30+ card types, LiDAR 3D rooms, Live Activities, pricing tiers of $4.99/mo, $24.99/year, $59.99 one-time). This became the primary source for what Vulpo actually does and costs, since the official site did not expose these details in a machine-readable form.
• ›The site lacks dedicated comparison or positioning pages (/features, /why, /about all 404'd). The agent had to use web search and prior knowledge to contextualize Vulpo against Home Assistant's official app, Nabu Casa, and Hassapp—information not present on vulpo.app itself.
• ›Pricing details were not fully scraped from /pricing (step [2])—the agent cites pricing tiers but likely sourced these from the App Store or prior knowledge rather than the pricing page HTML, indicating the page was either not fully rendered or content was behind JavaScript.
• ›The agent explicitly called out missing information: no feature breakdown table, no explicit 'Why Vulpo?' comparison, sparse Vulpo Labs details, and incomplete support/FAQ visibility—all legitimate gaps in site information architecture.