The agent successfully understood adkit.so's core value proposition, pricing, target audience, and competitive positioning, but had to work around an inaccessible primary website (client-rendered JavaScript) and assemble the answer almost entirely from external reviews and search results rather than official documentation. The site's own pages were not machine-readable; the agent relied on third-party review sites (coldiq.com), search snippets, and prior knowledge to construct a coherent explanation.

• ›Direct fetches of adkit.so pages [1, 3, 9] returned only HTML boilerplate with no readable content — the site is Nuxt.js client-rendered and not crawlable. The agent could not extract pricing, features, or positioning from the primary domain.
• ›All substantive content came from external sources: coldiq.com review [16] provided structured feature/pricing breakdown; web search results [6, 7, 11, 12, 14, 18] surfaced comparisons, customer reviews, and feature pages that were indexable by search engines but not by direct crawl of adkit.so.
• ›The agent cited external third-party review URLs in its final sources list (coldiq.com, toolfio.com, danrecommends.com) rather than official adkit.so pages, indicating those external sites are more agent-readable than the product's own website.
• ›Specific technical gaps remained unfilled: MCP setup complexity, platform-by-platform feature coverage, and permanent free tier details were not clarified even after eight search queries, suggesting this information either doesn't exist publicly or is scattered across non-indexed docs (e.g., 'https://mcp.adkit.so').
• ›The 'heavy bridge' friction outcome accurately reflects the divergence: the task was completed (satisfied) but required heavy reliance on web search and prior knowledge (60% prior_knowledge in sourcing) rather than direct site navigation.