Routines - Claude Code Blog

The /loop command runs a prompt or slash command on a repeating interval inside your active Claude Code session — polling a deploy, babysitting tests, iterating until something passes. Here's how it works, when to use it, and how it differs from routines.

Claude Code has six ways to extend it — CLAUDE.md, skills, subagents, MCP servers, hooks, and plugins — and they overlap just enough to be confusing. This decision guide explains what each one is, when to reach for it, when not to, and how to combine them.

Routines turn a saved Claude Code prompt into an unattended cloud agent that runs on a schedule, on a webhook, or on a GitHub event. Here is what they are good at, what they are not, and how to set up your first one.