ProductsOpen SourceBlog

Blog

What we learn while making things.

Notes on design systems, focused tools, and building for people and agents.

Your Figma plugin shouldn't need a new review to change.
Your Figma plugin shouldn't need a new review to change.
August 16, 2026

Figma's own Next.js example freezes the published plugin into a shell and moves the product into a web app you deploy. @xoxo-labs/figma-kit packages that pattern as a typed library with a React layer on top.

figma
figma-plugins
react
developer-experience
Your coding agent can't see its context meter. So we gave it one.
Your coding agent can't see its context meter. So we gave it one.
August 16, 2026

Context Sentinel uses Pi's Extension API to expose 80% and 90% context-window crossings to both the user and the model—without prescribing what the agent should do next.

pi
ai-agents
context-windows
developer-experience
Your AI agent doesn't need explanations. It needs rules.
Your AI agent doesn't need explanations. It needs rules.
August 16, 2026

How we cut our agent instructions by 44% — and made them easier to follow — by splitting every skill into a human rationale and a compact operational contract.

ai-agents
claude-code
developer-experience
tokens
Why Radix Colors?
Why Radix Colors?
March 31, 2023

A take on why Radix Colors is a better starting point for your colors.

design-system
radix-ui
color-system

Focused tools for people and AI agents. Less noise, better work.

ProductsOpen SourceBlog GitHub
© 2026 XOXO Labs.