compoundlab.ai
success
Claude Code
intent

What does compoundlab.ai do and who is it for? Explain it back to me.

7steps
31.1sduration
$0.2086cost
19,111tokens
9 steps2 reasoning steps
home
docs
home
/work
/services
docs
/contact
100%
answer from your site
57%
answer efficiency
67%
followed site links
answer sources
  • from this site100%
url discovery
  • given in the task14%
  • followed a link57%
  • guessed the URL29%
insight

Agent successfully gathered enough information to explain Compound Lab's core offering, audience, and differentiation, but pricing was completely absent from the published site and required direct contact. The site is well-structured for discovery but deliberately obscures commercial terms.

  • ›Pricing information was wholly absent: the agent attempted /pricing (404), checked /services (which described capabilities but no costs), and found only CTAs to book audits or contact sales. This gap is intentional site design, not a retrieval failure.
  • ›Core positioning came from homepage and /about fetches, both directly typed by the agent (not surfaced via navigation links). The site's own information architecture did not expose these as discoverable from a landing page—suggesting weak internal navigation or deep content.
  • ›/pricing and /contact both returned 404s, indicating either removed pages or incomplete URL structure. The site routes pricing inquiries through /start#audit instead, implying a deliberate gating strategy.
  • ›Agent filled gaps (case studies, project timelines, detailed methodology) by acknowledging them as missing rather than hallucinating—a sign the model correctly distinguished between site-derived facts and genuine absences.
  • ›The final response explicitly cites contact email (dylan@compoundlab.ai) and the /start#audit URL found during browsing, grounding those claims in fetched content rather than guesswork.

Help the next agent finish on compoundlab.ai

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 →