ora
writewise.com
success
Claude Code
intent

What does writewise.com do and who is it for? Explain it back to me.

8steps
42.3sduration
$0.2823cost
24,179tokens
10 steps2 reasoning steps1 search
home
docs
/how-it-works
home
/features
docs
/why-writewise
search
83%
answer from your site
29%
answer efficiency
67%
followed site links
answer sources
  • from this site83%
  • from search results17%
url discovery
  • given in the task14%
  • followed a link57%
  • guessed the URL29%
insight

The agent assembled a complete explanation of WriteWise from the homepage alone, but had to work around a broken site architecture (404s on logical navigation paths like /pricing, /features, /about) and fill gaps with external research. The homepage provided sufficient substance to answer what WriteWise does, who it targets, and how it differentiates, but pricing details remained opaque even after direct fetch attempts.

  • The homepage (steps [1] and [5]) was the only content-bearing source from writewise.com itself; all other expected navigation paths returned 404s, forcing reliance on model knowledge and external search results to complete the picture.
  • Pricing information was deliberately obscured on the site—no pricing page exists, and the homepage only hints at outcomes (15–20% decrements, 20% average savings) rather than publishing rates, leaving a critical task gap that the agent correctly flagged as 'unclear.'
  • The agent did not cite writewise.com URLs in its Sources section despite extensively fetching the domain; all three cited sources were third-party articles from external PBM market research sites (drugchannels.net, smithrx.com, wifitalents.com), indicating the external search results (step [8]) provided the competitive context and substantive details the site did not.
  • The agent accurately identified MavRx, the three plan tiers, and the self-funded employer target from homepage content, but technical depth on how MavRx works and concrete case studies were absent from the site, requiring the agent to note these as unfilled gaps rather than invent them.
  • Site navigability was poor: agent attempted 7 direct fetches/guesses (steps [1]–[7]), of which only 2 succeeded; the heavy reliance on 404s and external search (step [8]) to complete the task is reflected in the friction outcome 'succeeded_with_heavy_bridge.'

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 →