The agent gathered a coherent overview of Kindling Cannabis by combining the homepage fetch with targeted searches and supplementary third-party sources, but had to work around a JavaScript-heavy site architecture that obscured pricing, loyalty details, and product specifics behind postal-code-gated pages. The site itself provided core messaging (1-hour delivery, locations, product categories) but forced the agent to rely heavily on external sources (Leafly, Weedmaps, Curiocity, LinkedIn) and prior knowledge to complete the picture on pricing mechanics, competitors, and business model.

• ›Step [1] (homepage) and step [8] (cannabis-delivery page) returned hero messaging and high-level value proposition but rendered as heavily JavaScript-dependent HTML with truncated content, making full extraction difficult. Step [10] (locations) and step [18] (FAQ) similarly loaded but their rendered content was not fully visible in the response. Step [19] (Curiocity third-party article) was the only external fetch and appears to have been content-bearing for business model and operational details.
• ›Critical pricing and inventory information is location-gated behind postal-code entry on the live site—not fetchable statically. The agent could not access product pricing, specific minimum-order rules by location, or detailed menu from the site itself, and had to backfill from search snippets and third-party aggregators (Weedmaps, Leafly) and prior knowledge (franchise model, $200K-$250K capital requirement).
• ›The site lacks crawlable dedicated pages for 'About,' 'Pricing,' and 'FAQ' at predictable URLs (/about → 404, /pricing → 404, /faq → 404); instead, FAQ exists at /frequently-asked-questions. This forced the agent to discover URLs via web search rather than direct navigation, and to rely on external sources (LinkedIn, Leafly, RocketReach, Curiocity) for founder history, competitors, loyalty programs, and deal information that should reasonably be on-site.