ai-native-builder-workflow

This workflow enables non-technical individuals to build production-ready applications by orchestrating AI models as a technical co-founder, developer, and QA lead.

Core Principles

• The CTO Persona: Treat the AI as a Technical Owner. Instruct it to challenge your ideas, avoid "people-pleasing" (sycophancy), and own the technical architecture while you own the problem and user experience.
• Exposure Therapy: Gradually move from simple chat interfaces (ChatGPT/Claude Projects) to dedicated builders (Bolt/Lovable) to pro-level IDEs (Cursor/Claude Code).
• Planning over Vibe-ing: Never let the AI start coding until a markdown plan is finalized. Eager coding leads to architectural debt and complex bugs.

The /Command Workflow

Implement these custom prompts as reusable /commands within your AI coding environment (Cursor, Claude Code, or IDE system prompts).

1. Capture: /create-issue

Purpose: Quickly capture bugs or features without breaking development flow.

• Instruction: Tell the AI to stop what it's doing and summarize the thought into a specific format.
• Format: TLDR, Current State, Expected Outcome, and Priority.
• Integration: Use MCP (Model Context Protocol) to automatically create a ticket in Linear or GitHub.