The agent assembled a coherent explanation of Byway's business model, target audience, and positioning from web search results and third-party sources, but could not directly access the website itself (403/429 errors). This created gaps in pricing specificity, exact destination listings, and direct verification of claims.

• ›All substantive information came from search results [5, 10, 11, 13] and third-party articles, not from the target domain. Direct website fetches failed with access errors at steps [1, 3, 7, 8], forcing complete reliance on external sources.
• ›Pricing information was explicitly marked as non-public on the site itself ('not allowed to disclose individual prices'), yet the agent inferred payment structure and cost positioning from aggregated search snippets—a case where the site's opacity was surfaced by external coverage rather than its own pages.
• ›The agent correctly identified what it could not verify (specific destinations, customer demographics, current expansion status) and flagged these as missing rather than filling them with speculation, showing appropriate confidence calibration given access failure.
• ›Search results [5, 10, 11, 13] were rich enough to answer the core task (what, who, price model, differentiation), but the inability to fetch the website meant the agent could not validate positioning claims, see the booking UI, or confirm current offerings against historical information.
• ›No direct domain fetch succeeded; all routing through search results worked around the access blocks, leaving the agent dependent on how third parties and news outlets had summarized Byway rather than primary sources.