vaibhavt.com
success
Claude Code
intent
What does vaibhavt.com do and who is it for? Explain it back to me.
13steps
64.3sduration
$0.7459cost
134,777tokens
22 steps9 reasoning steps6 searches
home
docs
docs
docs
search
/work
search
search
docs
docs
search
search
search
100%
on-site discovery
71%
reliability
71%
link following
path origin
- previous resource71%
- prior knowledge29%
insight
The agent assembled a partial but reasonable understanding of vaibhavt.com by combining fragments from the homepage, blog landing page, and work section — but the site offered limited, client-side-rendered content that forced heavy reliance on prior knowledge and search inference. The agent correctly identified this as a personal developer portfolio with a blog, but could not find or confirm pricing, service offerings, contact methods, or detailed project descriptions, leaving key questions unresolved.
- ›Step [1] (homepage) and Step [5] (blog index) returned HTML that was heavily Next.js-rendered with truncated script bundles, making it impossible to extract actual page content from the raw responses alone. The agent had to infer structure and content from URLs and search results rather than from fetched markup.
- ›Step [9] (work page) and Step [15] (blog posts) similarly returned Next.js bundles with minimal readable content. The agent never actually retrieved legible descriptions of projects, services, or pricing—it inferred these from the URL structure and search result titles (e.g., 'Dynamic GIF Search | Vaibhav Tyagi' and 'Multi-Theme Magic: Next.js + shadcn/ui').
- ›The agent correctly noted that /about and /pricing returned 404s (steps [3], [4]), confirming those pages do not exist. However, the homepage itself did not surface any navigation to services, contact, or pricing, and the agent's HTTP responses contained no readable text to work with.
- ›The agent leaned heavily on prior knowledge (29% of fetches) and search results (steps [7], [11], [12], [17], [19], [20]) to build its narrative, rather than on content directly extracted from the site. Search result titles and snippets became the primary source of factual claims (e.g., 'Strava progress mirror for runners').
- ›The final response cites five URLs as sources, but only steps [1], [5], [9], and [15] were actually fetched during the run. The agent's answer draws heavily on inferences, search titles, and prior knowledge rather than on parseable content from those fetches.
Help the next agent finish on vaibhavt.com
Front builds the agent version of your site and shows which agents visit, what they came for, and what your pages couldn't answer.
or talk to us about agent readiness →