All activity
Arvind Ram Singh Kishoreleft a comment
This started from a recurring frustration: extracting a design system from an existing site is still manual and slow. Every migration or audit meant reverse engineering tokens and components from DevTools with no reusable output. The core problem is that design systems exist in production, but there’s no structured way to extract them.
extract-design-systemTurn any website into a reusable design system in seconds
Point it at a URL and it automatically extracts:
• Design tokens (colors, typography, spacing)
• UI components and patterns
• Reusable styles and structure
This is useful for:
• Designers auditing or rebuilding systems
• Engineers migrating or standardizing UI
• Teams documenting legacy frontends
• Rapid inspiration and competitive analysis
Built as a developer-first CLI and SDK, with outputs you can directly integrate into your workflow.
Open source and actively evolving.
extract-design-systemTurn any website into a reusable design system in seconds
Arvind Ram Singh Kishoreleft a comment
Inspired by YC's impact and AI agent tech like Deep Research. Solved fragmented startup research for jobs, investing, etc., by aggregating scattered data into AI-driven insights. Started with basic directory prototype; evolved to natural language search, then full autonomous agent with web crawling, cost controls, and real-time UI. Iterating on feedback.
YC AgentDeep Research agents for the YC Universe
YC Agent runs autonomous deep research on Y Combinator startups. It crawls public sources, analyzes founder signals, product positioning, funding history, and hiring patterns, and synthesizes everything into a structured intelligence brief — so candidates and investors can evaluate YC companies in minutes instead of hours.
YC AgentDeep Research agents for the YC Universe