The agent assembled a comprehensive overview of Currents by fetching the real currents.dev site (steps 8–9) after the preview domain's pages failed to render meaningful content. It successfully answered what Currents does, who it's for, and how it differs from alternatives, but could not retrieve exact pricing tier costs because the pricing page is client-side rendered and the static HTML did not include pricing figures.

• ›Steps [8] and [9] (currents.dev/pricing and currents.dev/) were the only content-bearing fetches; the preview domain (currents-a5npjnfyw-currents-a2561169.vercel.app) returned truncated HTML that provided no usable information, forcing the agent to pivot to the real site.
• ›The agent discovered the real domain via web search (step [5]) after the preview site's /docs and /features routes returned 404s. The preview appears to be a thin shell with no substantive content.
• ›Pricing information was incomplete: the agent found schema references to a $49/month base price and tier names (Scale, Business, Enterprise) but could not extract exact per-tier costs from the HTML. This is a client-side rendering problem—the page content is assembled in the browser, not in the initial HTML payload.
• ›The agent supplemented with web search results (steps [5], [6], [10], [11]) to build context on alternatives and orchestration features, indicating the currents.dev site itself does not present comparison matrices or full feature details in a static, fetchable form.
• ›The site is moderately agent-unfriendly for this task: core positioning and differentiators are documented, but pricing per tier and detailed feature comparisons require JavaScript execution or manual browsing.