Audit my entire database schema for design problems, beyond the basic user-content tables. List every table or collection, its fields, and its relationships, then review for the structural mistakes that cause pain at scale. Look for: god tables accumulating unrelated columns (a users table with 40 columns including billing state, preferences, and cached counts that belong elsewhere); fields storing two things at once (a status column mixing payment state and fulfillment state); relationships modeled as loose ID fields with no enforcement or documentation; arrays or JSON blobs being used where a real related table is the honest model (tags with their own metadata, line items with prices); and cached or computed values stored without a documented recomputation path (denormalized counts that drift when the recompute code forgets a case). For each finding, tell me the concrete problem it causes at 10x data volume or 10x features, and propose the fix as a migration plan, not a rewrite. Some findings will be acceptable trade-offs for a small app; say so explicitly rather than demanding purity. Don't change any schema in this pass. This is an audit and a plan. Deliver: the schema map, findings ranked by future pain, and the migration plan for the ones worth fixing now.