The agent assembled a reasonably complete explanation of Eden.so by combining sparse metadata from the site's own pages with external review content, but had to work around heavy reliance on client-side rendering and blocked third-party resources. The core positioning (research + ideation + creation + publishing for creators) emerged clearly, but specific feature-to-plan mappings and exact pricing tiers remained incomplete because the site's pricing page returned only metadata descriptions rather than rendered content.

• ›Eden.so's homepage, pricing, about, and help pages all returned Gatsby-compiled HTML with minimal rendered content—only meta tags and script references, no structured pricing tables or feature breakdowns. The agent extracted what it could from meta descriptions (e.g., 'file upload limits, storage, AI credits' mentioned in pricing page meta) but could not access the actual rendered UI.
• ›The agent successfully cited eden.so's own pages in its final response sources, but these pages returned no machine-readable feature matrices or JSON-LD pricing schema—only Open Graph metadata. The actual pricing details ($29, $79, $199) and plan names (Starter, Pro, Studio, Curator, Creator) came entirely from external review sites (creatoreconomytools.com, hypertools.so), not from Eden's own domain.
• ›The site is poorly agent-ready: JavaScript-heavy rendering blocked direct content extraction; third-party pricing resources (G2, Research.com) returned 403 Cloudflare blocks; and Eden's own help/documentation pages did not contain accessible pricing comparisons or feature details. The agent had to rely 50% on web search and external reviews rather than on Eden's own published content.
• ›The agent correctly identified its own gaps: it explicitly called out that exact feature-by-plan mappings, AI credit details, team collaboration specifics, and integration workflows could not be verified because the underlying pages were not fetchable. This demonstrates awareness of incomplete data rather than fabrication.
• ›The positioning vs. competitors (Notion, Ghost, Substack, Beehiiv) was inferred from context clues and review site framing rather than from Eden's own comparative messaging. Eden's site did not appear to publish a competitors page or explicit differentiation copy.