xerj.org
success
Claude Code
intent

What does xerj.org do and who is it for? Explain it back to me.

10steps
59.9sduration
$0.5038cost
106,279tokens
17 steps7 reasoning steps2 searches
home
/product.html
/use-cases.html
/for-agents.html
docs
search
/case-studies.html
/xerj-org/xerj
search
/playground.html
100%
on-site discovery
100%
reliability
88%
link following
path origin
  • previous resource88%
  • prior knowledge13%
insight

The agent successfully gathered enough information to explain XERJ's core function, target audience, and key differentiators, but encountered a significant gap: pricing and business model are genuinely absent from the site. The agent pieced together what XERJ does from the homepage, product page, use-cases page, for-agents page, case studies, and GitHub repo — all of which were directly fetchable and contained substantive information. However, the site does not publish pricing, deployment options, or commercial strategy in any machine-readable or human-visible form, forcing the agent to report this as a notable confusion point rather than an oversight.

  • ›Steps [1], [3], [4], [5] returned consistent, detailed positioning content (meta descriptions, page titles, hero copy) clearly stating XERJ is 'the search engine built for AI agents' with specific technical features (Elasticsearch-wire-compatible, Apache-2.0 licensed, zero-config indexing). The homepage meta description alone contained the core value prop and technical stack.
  • ›Step [10] (case studies page) and step [11] (GitHub repo) provided real-world proof points and benchmarks (5× fewer tokens, 2.7× vs grep, 26× vs memory alone) that substantiated the efficiency claims in the final answer.
  • ›The site explicitly publishes 'no pricing' — there is no pricing.html, no costs mentioned in any fetched page, and web search for site-specific pricing returned unrelated Xero accounting software results. This is not a navigation failure; it is a genuine absence. The agent correctly identified this and flagged the email-collection pattern on the playground as a signal of unexplained commercialization intent.
  • ›All key pages (product, use-cases, for-agents, case-studies) were linked from the homepage or discovered via navigation, making the site structure navigable. The agent did not need to guess URLs.
  • ›The agent correctly noted the 'V0.1' and April 2026 benchmarks as maturity signals, and the lack of comparison against Weaviate/Milvus/Qdrant as a positioning gap — these inferences went beyond what the site exposed but were reasonable given what was present.

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 →