How I think about agentic work, part two: graphs
How I structure agent work as a graph of plans, goals, tasks, and notes.
Writing
2026
How I think about agentic work, part one: loops
The loops I use to fix bugs, build features, and start projects with agents, and why I'd rather be on the loop than in it or out of it.
When AI writes all the code
I delegate implementation to agents while staying involved in design, evaluating larger features, and checking that the whole application works as I expect.
Running AI agents in devcontainers
How to use devcontainers to run AI agents like Claude Code and Codex in isolated environments you control, so you can skip the permission prompts.
The letter
New essays by email. It costs nothing.