The agent successfully gathered enough information to explain who Nico van Zyl is, what he does, and his positioning, but encountered significant gaps in pricing and engagement model information. The site is moderately agent-ready with structured data (JSON-LD, markdown exports, llms.txt) but lacks dedicated service, pricing, or contact pages—forcing the agent to assemble answers from biographical content and infer missing details.

• ›The agent's primary content sources were the about.md, career.md, profile.json, and llms-full.txt files—all explicitly cited and content-bearing. These provided 80% of the final answer (who, what, background, philosophy). Steps [3], [4], [5], [27] directly supplied the biographical, experience, and principle-based differentiation information.
• ›Pricing and engagement model information was entirely absent from the site. The agent discovered broken service pages (design-systems, progressive-web-apps, jamstack-websites all returned 404s), and web search results pointed to URLs that no longer exist, suggesting a recent redesign stripped away transactional details. The agent had to explicitly call this out as a gap rather than fabricate it.
• ›The site is well-structured for identity discovery (llms.txt, machine-readable JSON-LD, markdown exports with canonical URLs) but poorly structured for commercial intent. There is no contact form, no pricing page, no call-to-action, and no clear engagement model documented. The agent had to guess at common URLs (/contact, /work) rather than find them linked from the homepage.
• ›The site appears to have been recently redesigned to a minimalist portfolio model. Google's cache still indexes pages that now 404, indicating deliberate removal rather than natural site evolution. This mismatch between what search engines cache and what the live site serves created friction.