The agent could not fully access the portfolio site due to client-side rendering (Next.js) blocking static content retrieval, so it assembled an answer primarily from external sources: web search results, GitHub, LinkedIn, and other professional profiles. The site itself proved largely opaque to automated fetching—the agent had to work around inaccessibility by relying on third-party platforms and prior knowledge to piece together what Maarij Aqeel does, who he is, and what he's built.

• ›The homepage and common routes (/about, /projects, /api) all returned 404s or unrendered Next.js boilerplate with no static content, making the portfolio fundamentally unnavigable via HTTP fetch.
• ›The agent's successful answer came entirely from external sources: web search results [7][8] surfaced the portfolio's existence and Maarij's title ('AI Engineer & ML Developer'), while GitHub [9] and LinkedIn [10] provided project details, employment context, and professional history—none of which were accessible from the domain itself.
• ›The site exposed zero machine-readable metadata (no schema.org markup, no llms.txt, no public API) about its purpose, projects, or offerings; the agent had to infer everything from social/professional platforms and search engine snippets, indicating very poor agent-readiness for identity and positioning tasks.
• ›The agent correctly identified the site as a personal portfolio rather than a commercial product, but could only do so by pattern-matching the domain name against search results and external profiles—the site itself communicated nothing about itself.