Resources / Wild Eye AI

Wild Eye AI Labs

Practical tools. Working methods. Lessons worth keeping.

Labs is where we share resources developed through building, testing, and solving real problems with AI. Start with the AI Builder Playbook: a set of standards, prompts, and runbooks for keeping control of an AI-built application.

Featured release / Version 0.1.1

The AI Builder Playbook

Let AI build your prototype. Don’t let your prototype own you.

An app-building platform can get a prototype on screen quickly. The trouble starts when the next prompt rewrites the layout, invents a second styling system, or changes a file you thought was protected. You spend the afternoon fixing damage instead of building the product.

The AI Builder Playbook brings those lessons into a practical starting structure. Establish ownership before the first build prompt. Give each tool clear boundaries. Build the framework before the features. Check the work before accepting it.

Download the playbook Read on GitHub

Version 0.1.1 · ZIP download

Start with control. You hold exportable code, your repository, and the backend in your account where applicable. The AI app-building platform generates the prototype. Get the playbook at github.com/WildEyeAI/ai-builder-playbook.
View full-size graphic

01 / Inside the playbook

What you get

  • A kickoff runbook covering the first build, project boundaries, checks, and recovery points.
  • Reusable prompt blocks, including instructions for a prompt factory that turns requests into bounded tasks.
  • Governance, frontend coding, and design-system standards to adapt to your project.
  • A complete booklet with the explanations behind the rules.
  • Customization guidance and clear licensing.

02 / Audience

Who it’s for

Solo builders and small teams working with AI app-building platforms, often called vibe-coding platforms, where a prompt becomes a running prototype.

The focus is control of your code, repository, backend, and platform connections. It is not a general guide to coding agents or AI-assisted editors. A separate custom API is optional, and not every prototype needs a backend.

03 / Getting started

Start here

  1. Read the README and customization guide.

  2. Adapt the standards and establish ownership, permissions, and recovery access.

  3. Follow the kickoff runbook, inspect the generated foundation, and build one governed feature at a time.

These are working materials, not a certification or a substitute for testing and security review. The generalized workflow has not been exercised end to end across every supported setup.

By Henry V. Alfano, Founder, Wild Eye AI.

Licensing: CC BY 4.0 for Henry-authored prose and natural-language prompt templates; MIT for executable code samples. See the repository for the complete terms.