Claude Design System
An experimental design framework for building AI-assisted interfaces. Includes component primitives, design tokens, and automation recipes for rapid prototyping.
Creative Code Lab is a collection of experimental projects, tools, and research exploring the intersection of AI, creative code, and software development. We document our findings, share implementations, and push boundaries on what's possible with modern tooling.
An experimental design framework for building AI-assisted interfaces. Includes component primitives, design tokens, and automation recipes for rapid prototyping.
Interactive canvas for real-time code visualization and collaborative sketching. Combines live coding, visual debugging, and AI-assisted refactoring in a single environment.
Component library optimized for information-heavy interfaces. Tables, charts, dashboards, and grids designed for scanning, filtering, and interactive analysis at scale.
Experimental VSCode extension that uses Claude API for real-time code suggestions, test generation, and documentation. Context-aware and trained on project patterns.
How we approach systematic design at scale using AI feedback loops. Covers token generation, component automation, and the role of human review in quality gates.
Performance techniques for rendering thousands of rows without lag. Virtualization, memoization, and profiling strategies that scale to large datasets.
Deep dive into prompt engineering for code. Token budgets, context windows, and patterns that produce reliable, idiomatic output across languages.
Technical deep dive into WCAG compliance, perceptual uniformity in oklch, and how we built our navy + amber palette for accessibility and brand impact.
Creative Code Lab is a research group focused on exploring the future of interactive design and AI-assisted development. We believe that the best tools emerge when designers and engineers collaborate closely, and when experimentation is paired with rigorous thinking.
Our projects span creative coding, design systems, developer tooling, and human-AI collaboration. Each project is documented, open-sourced where possible, and designed to push our understanding of what's possible with modern technology.
Interested in contributing, collaborating, or learning more? Connect with us on GitHub, join our community Discord, or reach out directly.