Audit my database queries and add the indexes they need. Read through my actual query code (not just the schema): every find, filter, sort, join, and lookup my app runs. For each query pattern, identify the fields it filters and sorts on, then check whether an index covers it. Add indexes for: fields used in WHERE clauses on every request (user_id on owned records is the big one), fields used for sorting list pages (created_at, updated_at), unique lookups (email, slug, username), and foreign keys used in joins or lookups. Where a query filters on two fields together, create a compound index in the right order (equality fields first, then range/sort fields) rather than two separate ones. Do not index every field; each index costs write performance and storage, so justify each one by the query it serves. Do not add duplicate indexes that an existing compound index already covers. If my database supports it, show me the query plan (EXPLAIN) before and after for the slowest queries. Deliver a table: query, fields used, index added, and the expected improvement. Flag any query that can't be fixed with an index alone (like a text search doing substring matching) and note it for a separate fix.