The agent successfully extracted what LitFund does, who it serves, and its pricing model from the target site's homepage, then supplemented with external litigation financing context and specific case research to provide a complete picture. The site itself is minimally navigable—it exposed only core concepts without fee specifics, success rates, or case progress, forcing the agent to fill substantial gaps through external research and model knowledge.

• ›Steps [1] and [2] (both fetches of the homepage) returned the core business model, target audience, and three active case types, but the site disclosed fees only as a vague contingency phrase: 'Im Erfolgsfall wird LitFund am Ergebnis beteiligt' without specifying percentage. The agent noted this opacity in its final response.
• ›The site did not surface any comparison with competitors or explain what makes LitFund different operationally—the 'differentiation' section of the final response relied on model memory and inferences (bundling, cost transfer) from the site's positioning, not explicit differentiation claims.
• ›Steps [4] and [5] (external search results) provided critical context that was missing from the site: market precedent for ~35% contingency fees in Austria (step 4), specific insolvency details and investor counts for the Sun Contracting case (step 5), and competitor names like Padronus. These were cited in the response's sources but were not available from anlegerschaden.com itself.
• ›The agent had to type the base URL from memory twice ([1] and [2]) rather than discover it through navigation—indicating the site has minimal internal link structure or the agent did not follow any secondary pages. All three cases and fee structure came from the single homepage fetch.
• ›The final response correctly flagged the site's lack of case progress updates, success rates, and evaluation criteria as gaps—these were never present in the fetched content, making them genuinely missing rather than model-memory fill-ins.