The agent successfully compiled a comprehensive explanation of Fingerprint's purpose, pricing, target audience, and competitive positioning by assembling information from multiple sources. The docs.fingerprint.com site itself was poorly navigable for direct content retrieval (client-side rendered, 404s on guessed paths), forcing the agent to rely on the llms.txt artifact and prior knowledge/web search to build the answer. The agent worked around structural limitations by discovering and extracting the plaintext documentation index, then supplementing with external searches and general knowledge about the platform.

• ›Step [13] (llms.txt) was the primary on-site source, providing a structured index and overview describing Fingerprint as 'a device intelligence platform that helps you prevent fraud and improve user experiences' with coverage of SDKs, APIs, and use cases. This gave the agent its foundational framing.
• ›Step [16] (llms-full.txt) supplied detailed tutorial content on account sharing and account takeover use cases, demonstrating the platform's practical application areas and fraud prevention focus.
• ›Steps [3] and [4] (www.fingerprint.com homepage and pricing page) were fetched but returned client-side rendered HTML that the agent could not parse directly; the agent had to rely on prior knowledge and web search results to extract pricing tiers ($0 free, $99 Pro Plus, custom Enterprise) and feature details.
• ›The docs.fingerprint.com site published neither a human-readable overview page nor structured metadata (pricing, comparisons, feature matrix) that the agent could directly fetch. Guessed paths like /getting-started, /guides, /home all returned 404s, indicating poor discoverability for automated agents.
• ›The agent successfully identified llms.txt and llms-full.txt as artifact escape hatches when direct HTML fetches failed, demonstrating that Fingerprint has some LLM-friendly infrastructure but it requires knowledge of the artifact file convention.
• ›Pricing and competitive differentiation came primarily from web search results (steps [18], [19], [21], [22]) and prior knowledge, not from the docs site itself. The agent cited external comparison sites (SEON, G2, Capterra, TrustDecision) in its sources because the docs site did not provide this information.
• ›The agent explicitly flagged gaps: no exact Enterprise pricing breakdown found, no direct competitor comparison tables on the site, limited ROI/case study evidence, and unclear free plan enforcement rules. These gaps confirm the docs site prioritizes reference/tutorial content over sales/evaluation content.