Agent successfully gathered comprehensive information about Nothing Naughty despite the site blocking direct automated access; it pivoted to web search to discover entry URLs, then fetched the About page and product pages to assemble pricing, values, and positioning. The site's content was readily fetchable once discovered, but rate-limiting friction forced a workaround.

• ›Direct site access was blocked (HTTP 429), forcing reliance on web search results [5] to discover URLs that were then successfully fetched [7, 9, 11]. The agent's pivot was essential; the site itself did not surface these entry points via robots.txt or a sitemap.
• ›Product pricing and ranges were explicitly listed on fetched pages [9, 11], with complete AUD pricing for powders, bars, and sachets. This data was site-native and structured well enough to extract into a summary table.
• ›Company mission, origin story, and competitive positioning were explicitly stated on the About page [7], presented as narrative text. No structured data (schema.org, JSON-LD) was noted, requiring extraction by the agent.
• ›Flavor details and detailed nutritional information were identified as missing from the fetched product pages [11], indicating incomplete product-level documentation on the live site—a gap the agent correctly flagged rather than filling from memory.
• ›The agent correctly cited all four content-bearing fetches in its sources section, maintaining traceability. Search [5] supplied product category knowledge but was superseded by direct page fetches; it was not cited, correctly reflecting that the final answer drew primarily from site pages [7, 9, 11].