3 min read
cmux: A Terminal Built for Agentic Workflows
Why a terminal designed around running multiple AI agents simultaneously is worth paying attention to.
Agentic coding
Working with coding agents day to day: how to configure them, what they remember, and how to run several at once.
All posts · 5 posts
8 min read
Git Worktrees for Parallel Agent Work
Running multiple AI agents in parallel means running multiple branches in parallel. Git worktrees are how you do that without thrashing your working directory.
5 min read
OpenCode vs Claude Code: Same Idea, Different Philosophy
Both run AI agents in your terminal. Both support Claude. The differences are in architecture, openness, and what you can configure.
7 min read
Memory, Context, and Sessions in Agentic Coding
AI coding agents don't have memory. They have context windows. Understanding the difference changes how you work with them.
7 min read
Rules, Commands, and Skills: The Three Layers of Agentic Coding
Most developers use AI coding assistants as fancy autocomplete. The ones getting 10x more out of them understand the difference between rules, commands, and skills — and use all three deliberately.