The agent successfully assembled a comprehensive explanation of Gecko Security by combining fragments from heavily client-side rendered pages, blog posts, and web search results. The site's Next.js architecture made direct content extraction difficult, forcing reliance on meta tags, blog articles, and external citations; however, the agent found enough distributed information to answer all four core questions (what they do, who it's for, pricing, differentiation) and explicitly flagged gaps around exact pricing figures and language support limits.

• ›Meta tags on the homepage [1] and about page [4] contained the core positioning ('Security that actually understands your codebase'), but the HTML was heavily client-side rendered (Next.js), so most content detail was unavailable in the fetches themselves.
• ›The docs site [8] loaded successfully (200) but also appeared to be client-rendered, yielding no extractable text. Blog articles [17, 19] were similarly client-rendered and required external knowledge or search results to fill content gaps.
• ›Web search [16] surfaced direct blog URLs that clarified Gecko's differentiation (semantic analysis vs. AST-based SAST, business logic vulnerabilities, false positive rates), but these were blog pages the agent had already fetched—the search results' link titles and snippets likely contained more readable content than the full page fetches.
• ›Pricing details came from partial source material: free tier limits (3 repos, <20k LOC, Python/JS support) were inferred or cited, but exact dollar amounts for Professional/Enterprise tiers were explicitly called out as missing, indicating the site does not publicly disclose them.
• ›The agent cited four sources in its final response, but [17] (blog index) and [19] (single blog article) were both client-rendered Next.js pages that returned minimal extractable text—the agent's final answer likely drew more from search snippets [7, 16] and prior knowledge than from the full page HTML.
• ›The site's pricing page returned a 404 [3], forcing the agent to work around the missing direct resource and reconstruct pricing from fragments elsewhere.