Agent successfully assembled a comprehensive overview of Cashd by fetching the homepage, llms.txt machine-readable metadata, and the earn guides page. The site is moderately agent-ready: it publishes a detailed summary in llms.txt with structured feature and pricing sections, but pricing and business-model details are scattered or absent, and key claims (e.g., how 'earn' features work) are left unexplained on the public pages.

• ›llms.txt provided the most substantive, structured content: it explicitly listed features (photo scan, price guides, fee calculators, Resale Index), clarified that prices are asking-prices not sales data, and stated Cashd is free with no subscription tier. This file was discoverable from the homepage but not mentioned in marketing copy.
• ›Homepage and llms.txt agreed on core function but diverged on transparency: the homepage emphasized live pricing from StockX/eBay/TCGplayer without caveat; llms.txt revealed prices are estimates, not appraisals, and that Resale Index data mixes sources (deadstock vs. asking prices). Agent correctly surfaced this gap.
• ›The 'earn' page (trade-ins, cash back, gift cards, reselling guides) made claims without supporting detail: no partner names, rates, payout timing, or minimum thresholds. Agent noted this as a missing gap rather than falsely inferring content.
• ›Pricing was genuinely free for the core product (web + iPhone), but the site did not explain Cashd's revenue model or whether it takes a cut of sales. Agent correctly distinguished between marketplace seller fees (eBay, Poshmark, etc.) and any Cashd fee.
• ›Agent did not guess or insert model memory for unsubstantiated domain facts (e.g., competitor comparisons, accuracy metrics, account/privacy policies); instead it called these out as 'couldn't find,' demonstrating accurate attribution.