Cursor prompt

Optimize Database Queries

Identify N+1 queries, add missing indexes, rewrite slow queries, and add performance monitoring to an existing application.

Optimize the database queries in {{project_name}}.

## ORM: {{orm}} | Database: {{database}}
## Current pain points: {{pain_points}}

### Code to Audit:
```{{language}}
{{code_to_review}}
```

### Step 1: Identify Problems
Scan for:
- N+1 queries: a query inside a loop, or relations fetched without include/join
- SELECT * queries fetching columns that are never used
- Missing pagination — unbounded queries on large tables
- No indexes on WHERE/ORDER BY/JOIN columns
- Repeated identical queries in the same request (no request-level caching)
- Queries inside transactions that could be batched

### Step 2: Fix Each Issue

**N+1 Fix Patterns**
- Prisma: add include: { relation: true } to the parent query
- SQL: rewrite as JOIN or batch with WHERE id IN (array)
- Multi-level N+1: use DataLoader-style batching

**Index Additions**
- Single column index for equality filters
- Composite index for multi-column WHERE (order: equality first, then range)
- Partial index for low-cardinality filtered queries (WHERE status = 'active')

**Pagination**
- Replace OFFSET with cursor: WHERE id > $cursor ORDER BY id LIMIT $n
- Return nextCursor alongside results

### Step 3: Add Monitoring
- Log queries exceeding 100ms with full query text and bind params
- Add EXPLAIN ANALYZE to the 5 slowest identified queries

### Output
Refactored code, index migration SQL, monitoring setup, and before/after query count comparison.

Fill in: {{project_name}}, {{orm}}, {{database}}, {{pain_points}}, {{language}}, {{code_to_review}}. In Loopy you fill these in a form and copy the finished prompt in one click.