The agent assembled a reasonably complete picture of Tidmo's business model, target audiences, and value proposition, but had to rely heavily on external sources and prior knowledge to fill gaps left by the site's own content. The website itself is heavily JavaScript-rendered and does not expose detailed pricing, subscription tiers, or service details in a machine-readable or statically accessible form, forcing the agent to source most concrete information (pricing figures, competitor framing, business mission) from web search results and third-party coverage rather than the site's own pages.

• ›Step [1] (homepage) and steps [7, 8, 9] (key pages like /quem-somos, /saiba-mais, /quero-ser-uma-embaixadora) returned only JavaScript bundle boilerplate and minimal static content — the agent could not extract service details, pricing, or value proposition from the site's own HTML. The site is a Next.js SPA that does not pre-render content.
• ›Steps [5, 11, 12, 14] (web searches) were far more content-bearing than the direct site fetches: search results linked to Terra article framing Tidmo as 'Uber das diaristas,' a startup impact case study emphasizing formalization and professionalization, and Black Friday promotion pages. The agent cited all of these external sources in its final answer; the site itself did not surface this framing.
• ›Pricing figures (R$90 for 4 hours, R$148 for 8 hours) and earnings claims (up to R$2,500 monthly for workers) came from web search snippets and external coverage, not from the site's public pages. The agent correctly flagged this as outdated and subject to promotions, indicating the site does not expose current pricing transparently.
• ›The agent's differentiation analysis (social impact mission, 'Ambassador' branding, geographic focus to Northeast Brazil, professionalization of informal workers) was entirely sourced from external articles and the startup impact case study — not stated explicitly on the site's fetched pages.
• ›The site's navigation structure (URLs like /agendar, /quem-somos, /saiba-mais) is sensible, but the lack of static content rendering makes it opaque to agents that do not execute JavaScript. The agent had to search externally to learn what the site actually offers.