The week after WordPress 7.0 shipped, the dev chat agenda quietly listed the field guide, three dev notes, a PHP clarification and unicode email support. I run the same four-pass triage on every agenda like this before touching a client site. This is the checklist.
WordPress planned React 19 for 7.1, then moved it to a Gutenberg experiment. The bigger risk is plugins that bundle their own React, so check your builds and test a store on staging before core ships it.
Most advice on running agents at once is about attention: overview screens, alerts, recaps. On WordPress client work the shared checkout and the shared database break first. This is how I split the work, and when I stick to one session.
A chat model writes tidy solver code for five products and falls over on a real catalog. ORPilot interviews the user first and keeps raw data out of the prompt, and that pipeline is worth copying for AI features on a WooCommerce store.
WooCommerce 10.8.1 shipped on May 28 with fixes for a WooPayments onboarding crash and a fatal error during 10.7 to 10.8 upgrades. The fatal came from upgrade code running while the autoloader still pointed at the old plugin files, and the same guard works in your own plugins.
::checkmark styles the tick of native checkboxes, radios and selects without touching markup, which matters when WooCommerce owns the template. Here’s how I’d ship it, with a plain fallback first, @supports as the gate, and a plan for testing in Safari.
WordPress 7.0 shipped with a messy publish screen wherever the block editor isn’t in charge, and the fix is going out through the Hotfix and Classic Editor plugins instead of a core release. This is how to find the sites that need it.
An AI pass over your products can finish with valid JSON and no errors while it quietly corrupts the data. A better prompt won’t fix that. The queue WooCommerce already ships will: one item per action, retries, checkpoints, and output you can audit with a script.
contrast-color() reached Baseline in April 2026. On WordPress it matters because clients edit the palette, and hardcoded text colors on those presets keep breaking. One declaration covers most of it, with some limits.
An AI feature that works for three turns and then dies with a context length exceeded error has a design flaw. The transcript was never a database. Here’s where the run log belongs and how to keep prompt caching working.