Replit AI Tips & Prompts — Code Faster

Build a {{project_type}} application with the following specifications:

**Tech stack:** {{tech_stack}}
**Core features:**
{{feature_list}}

**Requirements:**
- Set up the project structure with clear separation of concerns
- Include a README with setup instructions
- Add environment variable handling (.env with example file)
- Set up the database schema and seed data
- Include basic error handling and loading states
- Make it mobile-responsive from the start

Start with the data model and core CRUD operations, then build the UI on top. Use sensible defaults for styling. Do NOT add authentication yet — I'll request that separately.

I'm getting this error in my Replit project:

**Error message:**
```
{{error_message}}
```

**File where it occurs:** {{file_path}}
**What I was trying to do:** {{intended_behavior}}
**What actually happens:** {{actual_behavior}}

Please:
1. Explain what's causing this error in plain language
2. Show me the exact fix (the specific lines to change)
3. Explain why the fix works
4. Tell me if there are related issues I should check for — this type of error often comes with other problems

Don't rewrite the entire file. Just show me the targeted changes needed.

My {{framework}} application on Replit is ready for deployment. Run through a pre-deployment checklist:

1. **Environment variables** — List all env vars the app needs and verify they're set in Replit Secrets
2. **Database** — Verify the database is configured for production (connection pooling, migrations applied)
3. **Security** — Check for: exposed API keys, CORS configuration, input validation, SQL injection risks, XSS vulnerabilities
4. **Performance** — Review for: N+1 queries, missing indexes, unoptimized images, bundle size
5. **Error handling** — Verify all API routes have proper error responses (not raw stack traces)
6. **SEO & Meta** — Check title tags, meta descriptions, Open Graph tags
7. **.replit config** — Verify the run command and ports are configured correctly

For each issue found, provide the exact fix. Prioritize by severity (critical > important > nice-to-have).

Set up a {{database_type}} database for my Replit project.

Application context: {{app_description}}

I need:
1. **Schema design** — Tables/collections with fields, types, relationships, and indexes. Include created_at/updated_at timestamps on all tables.
2. **Migration file** — SQL or ORM migration code to create the schema
3. **Seed data** — Realistic sample data (at least 5 records per table) for development
4. **Connection setup** — The database client configuration using Replit's database URL from environment variables
5. **CRUD helpers** — Basic create, read, update, delete functions for each main entity
6. **Type definitions** — TypeScript types/interfaces matching the schema

Use {{orm_or_driver}} for the database layer. Handle connection errors gracefully.

Create a REST API for {{resource_name}} in my {{framework}} project on Replit.

Data model:
{{data_model}}

Build these endpoints:
- GET /api/{{resource_name}} — List all (with pagination: page, limit, sort)
- GET /api/{{resource_name}}/:id — Get one by ID
- POST /api/{{resource_name}} — Create new (with input validation)
- PUT /api/{{resource_name}}/:id — Update existing (partial updates allowed)
- DELETE /api/{{resource_name}}/:id — Soft delete (set deleted_at, don't remove)

For each endpoint include:
- Input validation with clear error messages
- Proper HTTP status codes (200, 201, 400, 404, 500)
- Consistent response format: { success: boolean, data: ..., error?: string }
- TypeScript types for request and response bodies

Add a simple test you can run with curl to verify each endpoint works.

I'm working on {{feature_description}} in my Replit project and I want you to act as my pair programmer.

Current state of the code:
- {{current_state}}

What I want to achieve:
- {{goal}}

Rules for our session:
1. Before writing any code, briefly explain your approach (2-3 sentences max)
2. Write code in small, testable chunks — not entire files at once
3. After each chunk, tell me how to verify it works
4. If you see existing code that should be refactored to support the new feature, ask before changing it
5. Flag any architectural decisions that have long-term implications
6. Use the existing code style and patterns — don't introduce new conventions

Let's start. What's your approach and what's the first chunk of code?