Skip to main content
Organizations ship entirely new products with Blitzy that they would not have resourced otherwise. Instead of staffing a full team for months, you describe what you need — stack, architecture, features, user flows — and Blitzy generates a working codebase ready for review. Teams use this to launch internal tools, MVPs, and customer-facing products faster than traditional development allows.

Ready to build something new?

Use the New Product Template to get started with a structured prompt
Jump to: When to Use | Prompt Structure | Key Considerations | Common Patterns | Troubleshooting

​
When to Use This Approach

Build REST APIs, admin dashboards, internal utilities, and workflow tools that your team needs but hasn’t had bandwidth to build.

​
Prompt Structure

Follow the Golden Rules when writing your greenfield prompt. Focus especially on:
  • Vision & Purpose Clarity — Define the problem you’re solving, who the users are, and what makes this product valuable
  • Stack & Architecture Decisions — Specify tech stack, design patterns, and infrastructure choices upfront so Blitzy doesn’t guess
  • Core Requirements Completeness — List required behaviors starting with “System must…” or “Users must be able to…” to eliminate ambiguity
  • User Flow Specificity — Describe key journeys with entry points, steps, and success criteria so generated code reflects real workflows

​
Key Considerations

Start with a Clear Tech Stack

Specify frontend framework, backend language, database, and infrastructure choices explicitly. Leaving stack decisions open leads to inconsistent or unexpected technology selections.

Define MVP Scope Ruthlessly

Prioritize features into must-have vs. nice-to-have to keep the first generation focused. Trying to build everything at once produces shallow implementations across the board.

Specify Authentication & Access

User types, authentication methods, and access control rules need to be defined early. These decisions affect data models, API design, and frontend routing throughout the entire codebase.

Plan for Iteration

The first generation is a foundation, not a finished product. Structure your prompt knowing you will extend with follow-up generations using the Add Feature workflow.

​
Common Patterns

Full-stack web application — Define frontend and backend separately with clear API contracts. Specify data models, page layouts, and state management approach. Include authentication, error handling, and deployment configuration. API-first service — Focus on endpoint definitions, request/response schemas, data models, and integration contracts. Specify authentication, rate limiting, and error response formats upfront. Internal tool / admin dashboard — Emphasize workflows, data views, and role-based access. Define what data users need to see, what actions they can take, and how permissions map to UI elements.

​
Troubleshooting

Be more specific about stack choices and architectural patterns. Name the exact frameworks, libraries, and design patterns you want. Reference documentation or examples when possible.
Narrow your MVP scope. Use priority tiers (must-have, nice-to-have, future) and move non-essential features to follow-up generations using the Add Feature workflow.
Add explicit auth requirements to your prompt. Specify user types, authentication method (OAuth, JWT, session-based), and access control rules for each feature area.
Add the Require Test Coverage rule to your project on the platform and specify which test types (unit, integration, e2e) and coverage thresholds you expect.
Use the Add Feature workflow for subsequent iterations. Each follow-up generation builds on the existing codebase with scoped feature additions.