Hey Reader,

Open Design auto-detects 16 coding-agent CLIs and lets you swap engines with one click. The hard part is getting it running the first time.

This week opens with the 10-minute setup on Claude Code, plus two more reads behind it.

Open Design, running on your machine in 10 minutes

Open Design runs locally with Claude Code as the engine: install Node 24, enable Corepack, clone the repo, run pnpm tools-dev, open localhost, type your first brief. Ten minutes, start to finish.

The post is the full walk-through, including the one thing that trips almost everyone: the daemon scans your PATH when it starts, so if you launch the app from Spotlight instead of a terminal, it inherits a stripped-down PATH and reports your installed CLI as "not detected." The fix is in there, along with the bring-your-own-key fallback if you don't run a coding-agent CLI at all.

One honest note: this is the run-it-yourself path, built for people already comfortable in a terminal. If that's not you, the hosted route is still the better call.

Read the install guide →

Claude Code routines – work that runs while your laptop is closed

Most people automate Claude Code by bolting it onto a GitHub Action: hand-written YAML, secrets, and a prompt shelled out to claude -p. It works, barely. Routines replace that. They're saved configurations that run on Anthropic's cloud – define a prompt, connect your repos and MCP tools, pick a trigger (a schedule, an API call, a GitHub event), and Claude runs the job autonomously. No laptop awake, no YAML.

The difference that matters: a GitHub Action runs the commands you wrote. A routine receives a prompt and decides how to reach the outcome. The post is what they are, how to set one up, and where they actually fit.

Read the routines breakdown →

From sign-up to store in minutes – what automation actually buys a small business

A custom-merchandise company was drowning in copy-paste: every creator who signed up meant another product built by hand across Printify and Shopify – images, descriptions, variants, the lot. The fix was a pipeline of nine connected workflows that turned the whole thing into a push-button operation: 9,000+ creators processed, near-zero errors, sign-up to live storefront in minutes.

This is the case study, not a demo. It's the clearest picture we have of what production automation actually buys a small business – where the manual hours went, and what replaced them.

Read the case study →

Reply and tell me which one was the read for you. I read every reply.

– Alex

P.S. – The conversations behind posts like these happen in the WotAI Skool community. 760+ builders, three live calls a week, free to jump in.

Come build with us →