The agent successfully assembled a comprehensive explanation of AfterQuery by combining direct site fetches with external sources. The afterquery.com domain itself returned only HTML scaffolding without readable content—all substantive information came from Y Combinator and Forbes, which the agent discovered through web search and explicitly cited. The site is poorly navigable for automated discovery; standard routes like /pricing and /features returned 404s, forcing reliance on third-party coverage.

• ›Steps [0], [5], [11], [12] fetched afterquery.com directly but returned only Next.js boilerplate HTML with no readable text content. The site uses client-side rendering that leaves no indexable markup for the agent to parse.
• ›Steps [18] and [19] (Y Combinator and Forbes articles) were the only content-bearing fetches; they provided: business model (expert data curation for AI models), founders (Spencer Mateega, Carlos Georgescu), financial traction ($100M ARR, $300M valuation as of April 2026), professional network size (100k experts), and differentiation vs. competitors (quality-first, expert-validated vs. crowdsourced).
• ›Pricing information (custom contracts, $40–$250/hour expert rates) and specific product details came entirely from external sources, not from afterquery.com. The agent correctly noted that public pricing pages don't exist on the domain.
• ›Web search (steps [8], [9], [16]) was essential to bootstrap discovery—the agent guessed /pricing and /features, got 404s, then pivoted to external research. The 'succeeded_with_light_bridge' outcome reflects this: the task was satisfied but required leaving the domain.
• ›The site provides no machine-readable product summary, pricing schema, or structured data that would allow direct answering. The homepage exists but renders as empty in the agent's view, indicating either aggressive client-side-only rendering or a content gatekeeping pattern.