Agent successfully gathered a comprehensive explanation of Bundl's core offering, audience, and pricing model by fetching two pages (homepage and How We Work). The site publishes positioning and service structure clearly but withholds concrete pricing figures, equity terms, and technical definitions of 'AI-powered' work, leaving material gaps the agent correctly identified but could not fill from the site alone.

• ›Pricing is intentionally opaque: the site describes a monthly commitment model with three tiers and risk-sharing options but provides zero numbers, ranges, or examples. The agent correctly called this out as 'confusing or not found' rather than inferring from model memory.
• ›The Launchpad entity is explained structurally (test, sell, contract, decide) but governance and post-launch ownership are not addressed. The agent identified this as a legitimate gap rather than guessing.
• ›Both pages were discovered without site navigation: step [0] was typed from memory (not surfaced by internal links), and step [1] was a plausible URL guess that succeeded. The site's own navigation did not guide the agent to the How We Work page; the agent inferred it existed.
• ›Impact claims (20M+ consumers, €2B+ value, 80% repeat clients) appear on the homepage but carry no source, methodology, or date. The agent noted this lack of grounding without inventing credentials.
• ›The agent did not fetch Cases, Knowledge Hub, or Venture Club pages despite linking to them, and acknowledged this scope limit transparently. The final answer is grounded only in what was actually retrieved, not speculated about.