The agent gathered substantial information about Nocturne's purpose, target audience, and differentiation from Nightscout by consulting external documentation (GitBook FAQ) and GitHub repositories, but could not locate public pricing for Nocturne Cloud despite visiting the main site and conducting multiple targeted searches. The getnocturne.dev site itself proved difficult to navigate due to client-side rendering that obscured content, forcing the agent to rely heavily on off-site sources and prior knowledge to construct a coherent answer.

• ›Step [11] (GitBook FAQ) and step [15] (GitHub search results) provided the core product narrative—that Nocturne is a modern Nightscout successor with WebSocket-first architecture, native device integrations, and better security. This critical content was NOT found on getnocturne.dev itself.
• ›The main site (steps [1], [3], [4], [5], [7]) returned only HTML scaffolding with no rendered content visible in the fetches, indicating heavy client-side rendering. Step [4] confirmed /pricing returns 404, and no pricing page exists at getnocturne.dev.
• ›The agent was forced to assemble the answer from fragmented sources: product positioning from GitHub docs, feature details from GitBook, and comparative context from web search results. The site's own pages did not expose pricing, feature comparisons, or audience information in a machine-readable or easily discoverable form.
• ›Step [14] (GitHub usenocturne/nocturne repository) and step [15] (GitHub nightscout/nocturne repository) were cited as sources but returned HTML scaffolding without visible content in the fetch response, suggesting the agent relied on URL metadata and prior knowledge rather than parsed content.
• ›Pricing remains a gap: neither self-hosted tier descriptions nor Nocturne Cloud's commercial model are published on the site, making the billing-commerce layer opaque to agents and likely confusing to humans.