The Pre-Build Interrogation Prompt

Use it when you have an idea for a feature, bug fix, or product change, but you want the agent to inspect the project, clarify risks, and create a plan before touching the code.

Jun 26, 2026

∙ Paid

The Pre-Build Interrogation Prompt is a companion resource to my article “Are You Sure You’re Actually Vibe Coding?”

Read the article first because this prompt only makes sense once you understand the difference between vibe coding, slop coding, spec-assisted building, and spec-driven development.

What The Pre-Build Interrogation Prompt Is

This prompt helps you apply that framework before a coding agent touches your project.

Use it when you you’re not yet sure whether the AI should start coding. Instead of asking the agent to implement immediately, you ask it to interrogate the idea first.

Use it before:

Building a new feature
Refactoring an existing flow
Changing auth, payments, permissions, or data handling
Asking an agent to edit multiple files
Turning a vibe-coded prototype into something users will actually touch
Moving from exploration into a more structured build mode

The prompt makes the agent inspect the existing project context, identify unclear requirements, surface hidden assumptions, map user roles and permissions, flag data model impact, find edge cases, assess security and privacy risks, estimate blast radius, define shippability criteria, and recommend tests before implementation.

In plain English: it stops the AI from confidently building the wrong thing.