docs.sqd.dev
success
Claude Code
intent

What does docs.sqd.dev do and who is it for? Explain it back to me.

14steps
55.8sduration
$0.6482cost
137,134tokens
15 steps1 reasoning step3 searches
home
search
/overview
llms.txt
/en/sdk/subsquid-vs-th…
/sdk
/network
home
docs
search
/en/portal/pricing
llms-full.txt
search
/compare
64%
on-site discovery
100%
reliability
45%
link following
path origin
  • previous resource45%
  • web search36%
  • prior knowledge9%
  • other9%
insight

The agent successfully gathered comprehensive information about SQD's purpose, target audience, pricing model, and competitive differentiation by assembling fragments from multiple fetches and searches. The site is reasonably navigable for structured discovery (docs exist, comparison pages are published), but pricing information is incomplete (Portal API pricing marked "coming soon"), terminology is fragmented across multiple product names, and the relationship between products requires inference. The agent filled gaps through web searches and prior knowledge rather than relying on a single authoritative explanation.

  • ›Step [6] (llms.txt) and [11] (llms-full.txt) provided machine-readable documentation indexes, enabling the agent to discover and navigate to specific pages (e.g., pricing, comparisons), demonstrating good documentation structure for agents.
  • ›Steps [9] and [12] surfaced detailed competitive comparisons (SQD vs The Graph) that directly answered the 'how is it different' portion of the task, showing the site publishes this content explicitly and at fetchable URLs.
  • ›Pricing information was fragmented: step [8] (docs.sqd.dev/en/portal/pricing) returned a page marked as 'coming soon' for Portal API, while SQD Cloud pricing was found only through external navigation (step [5] to sqd.dev/pricing). The agent had to assemble pricing from multiple sources and acknowledge incompleteness.
  • ›Product terminology is scattered across the site (Portal API, Squid SDK, Pipes SDK, SQD Cloud, SQD Network) with no single page clarifying which tool solves which problem—the agent had to infer relationships from separate documentation pages rather than from a unified positioning page.
  • ›Steps [2], [4], [10], [13] (main domain sqd.dev pages) returned HTML without text content in the trajectory (truncated responses), limiting their utility; the agent relied instead on doc-site pages and searches to extract substantive answers.
  • ›Web searches (steps [1], [7], [12]) were necessary to locate comparison pages and confirm pricing details, suggesting the site's navigation does not lead organically to comparative positioning or all pricing tiers from the docs homepage.

Help the next agent finish on docs.sqd.dev

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 →