WordPress Core AI: the MCP adapter goes standalone
Where the WordPress Core AI work stands before 7.0: the MCP adapter moving to a standalone plugin, the connector approval experiment, and a release plan that puts polish ahead of new features.
Where the WordPress Core AI work stands before 7.0: the MCP adapter moving to a standalone plugin, the connector approval experiment, and a release plan that puts polish ahead of new features.
RAG re-derives everything on every query, so nothing accumulates. This is the Raw and Wiki folder layout, the three control files, and the daily, weekly and monthly jobs I use to keep an AI context layer my agents can read.
WordPress 7.0 slipped to May 20, 2026 because RC3 is closer to a second Beta 1 than a release candidate. Early real-time collaboration code fired around 4,000 SQL queries in two minutes of editing, so core is rebuilding the storage layer.
WordPress 7.0 (Armstrong) is past the foundational AI work and into ecosystem plumbing. The connector approval experiment blocks plugins from your API keys until an admin says otherwise, and RC1 lands May 14, so provider plugins need updating now.
Real time collaboration is out of WordPress 7.0, and the benchmark data says why. Across eight hosting environments, custom-table-with-transients came out about 52% faster than the post-meta baseline, and post-meta-transients collapsed on hosts with no object cache.
WordPress 7.0 RC3 landed without Real-Time Collaboration. Core pulled it over race conditions, memory use and server load, and pushed the work to 7.1. This post covers testing RC3 from WP-CLI and what plugin authors built on the RTC endpoints need to change.
WordPress.com opened the AI Assistant to every paid plan and shipped Studio Code, a beta CLI agent for local work. Notes on what the CLI does with block markup and what the plan change means day to day.
Lists make you pay an O(n) memory shift on every pop(0). This walks through collections.deque instead: O(1) at both ends, maxlen for fixed-size history buffers, and a rolling average example.
The WooCommerce REST API documentation left its 4.2MB single-page site for per-endpoint files on developer.woocommerce.com. What that changes for coding assistants, raw markdown fetches, and doc updates that ship inside the PR.
Dynamic typing is fine in a notebook and expensive in a long-running pipeline. Notes on using type annotations, TypedDict, Literal and union types so a bad value fails at check time instead of four hours into a run.