The agent retrieved core product information but had to navigate around poor information architecture to find pricing—the /pricing page exists but is not linked from the homepage or docs page, forcing the agent to guess the URL. Missing details on code host support, security, data handling, and integrations prevent a complete picture.

• ›Pricing is hidden from discovery: The /pricing page was not linked from the homepage or /mcp page and was only reached because the agent typed the URL from memory. For a product with a pricing tier, this is a critical discoverability failure for both human and AI visitors.
• ›Security and data handling are absent: Neither the homepage nor /mcp page discloses how code is accessed, stored, protected, or how customer conversations are retained—information material to adoption decisions.
• ›Code host and language support are undocumented: The site does not state which platforms (GitHub, GitLab, etc.) or languages are supported, forcing customers to email for basic compatibility information.
• ›AI client integration is jargon-heavy: The /mcp page uses technical terms (MCP, WebMCP, llms.txt, robots.txt, Accept header) without explanation, making it hard for non-technical readers to understand what 'AI-client ready' actually means in practice.
• ›Site structure hides key information: Three separate pages (homepage, /mcp, /pricing) are required to answer the user's questions, but only the homepage is discoverable from the domain root—the other two require external knowledge or guessing.