verify-then-draw

Verify code correctness before rendering architecture diagrams from modules.

15|6|Updated May 7, 2026
One-click install
npx skills add https://github.com/MinhThang1009/dotclaude --skill verify-then-draw
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: verify-then-draw
Source: https://github.com/MinhThang1009/dotclaude/tree/main/plugins/verify-then-draw/skills/verify-then-draw
Command: npx skills add https://github.com/MinhThang1009/dotclaude --skill verify-then-draw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Audit and render diagrams from codebases without trusting buggy code, preventing misrepresentation in documentation.

Core Features & Use Cases

  • Enforces a three-tier gate (T0, T1, T2) ensuring code correctness, diagram accuracy, and notation readability before rendering.
  • Provides a project-agnostic workflow with invariants, denominator route/state counting, and a human-in-the-loop gates for safety.
  • Suitable for architecture diagrams (use-case, sequence, state, ERD, deployment, pipeline) derived from verified code in software projects.

Quick Start

Run the three-tier gate workflow on your module to generate diagrams only after code verification and human gates pass.

Frequently Asked Questions about verify-then-draw

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate accurate architecture diagrams from source code without rendering buggy logic?

To generate accurate architecture diagrams from source code without rendering buggy logic, you enforce a three-tier verification gate that checks code correctness, diagram accuracy, and notation readability before producing any visual output.

What is the best way to audit code before creating a PlantUML sequence or deployment diagram?

The best way to audit code before creating a PlantUML sequence or deployment diagram is applying a gated workflow that requires explicit invariants, live execution data, and denominator topology to verify the underlying code first.

Can I use a verification gate to generate use-case, ERD, and state diagrams across different modules?

Yes, you can use a project-agnostic gated workflow to generate use-case, ERD, state, and component diagrams across different modules by verifying code correctness through route counting and AST topology before rendering.

Why does my automated UML diagram misrepresent the actual codebase architecture?

Automated UML diagrams misrepresent the actual codebase architecture because they trust buggy code without enforcing explicit invariants, live execution data, or a human-in-the-loop safety gate during the rendering process.

How to ensure notation readability when drawing component diagrams from verified code?

To ensure notation readability when drawing component diagrams from verified code, you apply a tiered gate checking code correctness, diagram accuracy, and notation readability sequentially before final rendering.

Do I need live code execution data to safely render pipeline diagrams from my codebase?

Yes, you need live code execution data and explicit invariants to safely render pipeline diagrams, ensuring the visual output faithfully reflects the underlying code topology through a human-in-the-loop verification gate.