Run routine business work from the terminal

Use Claude Code as the execution layer for recurring business tasks while keeping a human check before anything risky.

Done with Claude Code· Advanced· Apr 2026

What they did
A poster swapped the Claude web app for Claude Code in the terminal and treated it as the part of the setup that carries out business operations. They described Claude Code as the business’s infrastructure, while Claude.ai stayed in place for creative and design work. The two tools were linked, with Claude Code handling execution and Claude.ai handling more open-ended thinking.
How it went
They say the business now runs through Claude Code instead of Claude.ai. The thread did not include task-by-task results or business metrics, so the clearest takeaway is the workflow change rather than a measurable gain. Commenters flagged safety limits, especially around broad permissions, command review, and rollback.
Tips
Keep a human in the loop for commands, approvals, and recovery steps. Use Claude.ai for creative or design tasks if you want to separate ideation from execution.

Try it yourself with Claude Code

I run a [type of business]. Create a terminal-based workflow with a project folder that includes a short business overview, key customers, tools, and recurring jobs. Then carry out these [2-3 specific tasks], saving each output in a separate file and ending with a brief recap of what changed. Before you send anything, pay money, remove files, or submit forms, stop and ask for my approval first.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Read the original on reddit.com ↗

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.