The agent successfully understood jishnuteegala.com by fetching markdown versions of the site's content pages. The site is a personal portfolio for a DevOps Platforms Engineer, well-structured and agent-ready through its explicit markdown export feature (appending .md to URLs). The agent assembled a complete answer covering what Jishnu does, who the site targets, that there is no pricing model, and how it differentiates—all grounded in fetched content.

• ›Steps [7–10] returned markdown versions of all key pages (index, CV, projects, labs) that contained the substantive information about Jishnu's role, skills, employment history, open-source projects, and site structure.
• ›The site explicitly advertises markdown availability via a comment in the HTML ('Full site index for agents: https://jishnuteegala.com/llms.txt — every content page is available as markdown by appending .md'), making it highly discoverable and agent-optimized; the agent recognized this pattern and pivoted to markdown fetches.
• ›The agent did not need to guess or rely on prior knowledge to construct the answer—all facts (UBS employment, tech stack, project names, CV details) came directly from fetched markdown content; prior knowledge was minimal (13%) and used only for initial navigation.
• ›The Labs section was genuinely empty ('Nothing here yet'), which the agent correctly reported as a gap rather than attempting to invent content.
• ›No pricing information exists because this is a personal portfolio, not a commercial product; the agent correctly identified and articulated this distinction.