I’m curious how some of you are using Local with AI tools and what kinds of workflows you are doing.
For me, I’ve been starting to use Claude Code for various things related to building Local (ie, the Electron app) but over the last few weeks I’ve started using it to help QA various Genesis projects in preparation for the WP 7.0 release.
Honestly, the “Claude Code + Local + WP Site” workflow wasn’t as seamless as I’d like it to be, so I’d like to make it better and would love to hear how we can make it better for you too!
I’ll start by articulating my experience and maybe others can chime in:
What I wanted to do was start Claude Code and type in a prompt along the lines of:
I’m doing some QA for a Genesis theme as outlined in the Jira issue LOC-1234. Can you create a new QA site with the latest WP Beta release and symlink the various repos I have already cloned down in
~/src/genesis
… and then “poof” I could poke around in the WP admin, chat with Claude to visit and troubleshoot various pages using playwright-cli, import demo content, etc
But the actual process was a bit more manual and had a few things I needed to manually adjust.
Most of the hurdles were around
- Not having a great way to create and manage specific site settings from the agent I’m chatting with
- Needing to manually write up the context for how an AI agent orients itself within a WP site managed by Local (ie, "This is a WP site in Local by WP Engine. The url is … you can use wpcli by setting these env vars…)
So, anyway, I’d love to hear how you’re using AI+Local+WP – what pain points are you encountering and what do you wish could be improved in the workflow?