The agent assembled a reasonably complete understanding of ChatEDU's purpose, audience, and differentiation, but could not access pricing details or confirm exact feature tiers—working around an inaccessible main website by relying on GitHub repositories, Devpost, and web search snippets. The site's heavy JavaScript rendering and missing /pricing endpoint forced heavy reliance on external sources and prior knowledge rather than primary product pages.

• ›Step [8] (Devpost) and steps [11], [15] (search results linking to GitHub) provided the most substantive product information—describing the LMS + AI tutoring hybrid, teacher dashboards, knowledge graphs, and multi-agent framework. These were not sourced from chatedu.io itself but from GitHub and hackathon documentation.
• ›The main chatedu.io homepage ([1], [7]) returned only HTML skeleton with no readable content due to heavy client-side rendering, and both /pricing and /en/pricing routes returned 404s ([3], [4]). The agent could not access the official pricing page and had to infer pricing ($3–$8/month) from search results, without confirming tier details or feature mapping.
• ›The site is not agent-ready for discovery tasks: no structured metadata in homepage HTML, no accessible pricing endpoint, no llms.txt or API documentation on the main domain. The agent had to pivot to GitHub (a third-party platform) and Devpost to reconstruct the product story, indicating the company's marketing site does not self-serve machine-readable product information.