What does riccardoerra.com do and who is it for? Explain it back to me.
- previous resource50%
- prior knowledge50%
The agent successfully assembled a comprehensive overview of riccardoerra.com by fetching the homepage and a year-in-review page, both of which returned rich HTML metadata and content. The site was highly navigable and machine-readable — the meta description alone answered most of the task (what Riccardo does, his role, his focus areas), and the homepage HTML contained sufficient semantic structure to extract information about his identity and professional positioning. However, the agent had to rely on prior knowledge about Figma employment and project details that were not fully explicit in the fetched responses, and it correctly identified gaps (no pricing, no detailed case studies, no clear service offerings).
- ›Step [1] (homepage) returned a well-formed meta description tag that directly answered 'what they do' and 'who it's for': 'Riccardo Erra is a developer advocate based in London. He thinks a lot about design systems and AI.' This is the kind of agent-ready semantic signal that grounds the entire answer.
- ›The homepage HTML included explicit keywords listing his roles (front end developer, freelancer, contractor, designer) and geographic/social context, making his professional identity unambiguous without requiring further interpretation.
- ›Step [3] (2024 year in review) confirmed the site's minimal content model — only a handful of pages, no blog or articles section — which directly informed the agent's finding that the site is a 'business card' rather than a comprehensive portfolio.
- ›The agent correctly identified what was NOT present: no pricing information, no service offerings, no detailed project case studies. This gap analysis was accurate and required no guessing — the site genuinely does not expose these details in fetched form.
- ›The agent drew on prior knowledge (Figma employment, Mews design system role) that was not present in the two fetches; however, these were positioned as contextual amplification rather than core findings, and the agent explicitly listed what it could NOT find.
- ›The site's extreme minimalism — static HTML, no interactive elements, no hidden forms or modals — made it trivially navigable for a web agent; there were no discovery barriers, API endpoints, authentication walls, or obfuscated content (aside from the cloudflare-protected email).
Want to run your own?
Join the waitlist for early access to point your own agents at any domain, with the intents you choose.
or talk to us about agent readiness →