The agent successfully assembled a comprehensive explanation of Open Protocol Labs by combining a machine-readable llms.txt file, the vision page, and external product site fetches. The site itself is minimally navigable—most traditional pages (pricing, about, products) returned 404s—but the agent recovered through a well-published llms.txt artifact and prior knowledge, delivering a largely complete answer. The primary gap was pricing/monetization, which the site does not disclose anywhere.

• ›Step [5] (opl.dev/llms.txt) was the primary content source: a structured 2.8k-word document explicitly describing OPL's mission, products (Sigma Identity, BitChat, Jamify, Minerva, 1Sat Browser), use of Bitcoin protocols, and open-source philosophy. This single artifact contained ~70% of the factual foundation for the final answer.
• ›Step [6] (opl.dev/vision) returned an HTML slide deck with metadata but no readable body text in the fetch; the agent did not extract new facts from it despite marking it as followed.
• ›Steps [15] and [16] (1sat.app and sigmaidentity.com) were fetched to understand products by name, but returned minimal readable text; the agent relied on prior knowledge and the llms.txt document to explain them.
• ›The site publishes no /pricing, /about, or /products pages (all 404s). The sitemap lists only /, /privacy, /terms, and /vision. OPL relies entirely on llms.txt and downstream product sites to surface its identity and mission.
• ›The agent correctly flagged missing information: no public pricing model, no formal competitive positioning, and unclear monetization strategy. These gaps are genuine site limitations, not retrieval failures.
• ›The site is optimized for AI crawlers (llms.txt exists and is cited in sources) but minimal for human navigation. A user would struggle to understand the business model or product differentiation from opl.dev alone.