The agent successfully assembled a comprehensive explanation of Bank of Georgia (sb.ge) by pivoting away from the primary domain after discovering it was a JavaScript-heavy SPA with no readable HTML content. It relied entirely on web search results and cited sources to answer all four questions (what they do, who it's for, pricing, differentiation), working around the site's fundamental inaccessibility to crawlers. The primary domain sb.ge was not agent-navigable; all substantive answers came from external sources.

• ›Steps [1], [3], [4], [5], [9], [12], [13] fetched from sb.ge and bankofgeorgia.ge but returned only script imports and meta tags—no readable content. The agent correctly identified this as a JavaScript SPA and pivoted to search.
• ›All substantive content came from web search steps [7], [10], [15], [16], [18], which returned links and snippets about Bank of Georgia's services, competitors, and positioning. The agent explicitly cited five sources in its final response, all discovered via search.
• ›The agent had to assemble the answer from fragments across multiple search queries and external sources (Wikipedia, CBInsights, thebanks.eu, Global Finance awards, hashbank). No single fetched page contained the full picture.
• ›Pricing information was partially answered (ATM fees, conversion fees) but the agent clearly stated 'difficult to extract' and 'couldn't find clearly' for interest rates, maintenance fees, and service tiers. This gap was acknowledged and transparent.
• ›The site itself published no machine-readable identity markers (structured data, llms.txt, API docs, or readable HTML) that would help an agent understand the business model without external search.