MECE risk framework for AI-generated code — 5 dimensions, 4 tiers, actionable mitigations.
An interactive React component that classifies vibe-coding risk across 5 orthogonal dimensions and maps results to concrete mitigation tiers. Accompanied by full AsciiDoc documentation with 30+ verified references.
-
5 MECE Dimensions: Code type, language safety, deployment context, data sensitivity, blast radius
-
4 Cumulative Tiers: Each tier includes all mitigations of lower tiers
-
Radar Chart: SVG visualization of the risk profile
-
Presets: Quick scenarios (Landing Page → Payment Service)
-
DE / EN: Full bilingual support, switchable at runtime
-
Inline Documentation: Slide-out sidebar with all 6 doc sections
-
Mobile-friendly: Responsive layout with stacked sliders
The included GitHub Action (.github/workflows/deploy.yml) handles everything automatically:
-
Builds the React app with Vite
-
Renders the AsciiDoc documentation with Asciidoctor
-
Deploys both to GitHub Pages
-
Go to Settings → Pages in your GitHub repository
-
Under Build and deployment, select GitHub Actions as the source
-
Push to
main— the workflow runs automatically
After deployment:
.
├── .github/workflows/deploy.yml # GitHub Pages CI/CD
├── docs/
│ └── risk-radar.adoc # Full AsciiDoc documentation (30+ refs)
├── src/
│ ├── RiskRadar.jsx # Main React component (self-contained)
│ ├── main.jsx # React entry point
│ └── index.css # Minimal reset
├── index.html # Vite entry HTML
├── package.json
├── vite.config.js
└── README.adocThe docs/risk-radar.adoc file contains the full framework documentation:
-
Empirical evidence (Veracode, CodeRabbit, METR, Perry et al.)
-
5×5 dimension matrix with detailed descriptions
-
4-tier mitigation model with deterministic, probabilistic, and organizational measures
-
Regulatory context (EU AI Act, HIPAA, PCI-DSS, IEC 61508, DO-178C, ISO 26262)
-
References: 30+ verified URLs to primary sources