The agent gathered a broad understanding of RICHARK's business model, target audience, and core pricing through web search and blog access, but could not fully satisfy the task due to Cloudflare blocking the main website and membership portal. The agent assembled a coherent picture of what RICHARK does (real estate education + integrated advisory services) and who it's for (aspiring landlords, financial learners), but exact current membership pricing and competitive differentiation remain incomplete because the primary domain and pricing pages were inaccessible.

• ›The blog subdomain (blog.richark.com.tw, step [6]) was the only directly fetchable RICHARK property and became the primary content source; it provided course mentions and general positioning but lacked pricing details.
• ›Web search snippets and third-party blog reviews (steps [9], [12], [13], [15], [16]) were the agent's main source for pricing data (100,000 NTD tutoring fee, 1,000 NTD consultation, course examples at 2,980–3,980 NTD), none of which came from RICHARK's own infrastructure.
• ›The main domain (www.richark.com.tw), membership portal (member.richark.com.tw), and shop (shop.richark.com.tw) all returned Cloudflare 403 responses, making it impossible to verify current pricing, full course catalogs, or membership tiers directly from the source.
• ›The agent relied on prior knowledge (~13%) and reconstructed information from cached reviews and third-party sites to fill gaps; the site does not publish machine-readable pricing or positioning in an openly crawlable layer.
• ›Competitive differentiation had to be inferred from review content and general positioning rather than from explicit RICHARK statements; the agent could not access primary comparative materials.
• ›The blog is agent-accessible but is primarily a content/marketing channel, not a transactional or comprehensive information hub—courses, memberships, and detailed offerings are gated behind the blocked portals.