> [!question] Lingering Questions > - What would change this claim if it turned out to be wrong, and where's the strongest counter-evidence? > [!idea] Article Ideas > - The real inflection point (L2→L3) is a mentality shift from pointing agents at static files to letting them touch live staging/production sites. > [!info] Cross Reference > - See MOC links below for the cluster this note belongs to. > [!abstract] Vocabulary > - (no source-specific jargon requiring definition in this piece) --- ## Metadata - **Source:** [[3. Resources/Readwise/Articles/The 5 Levels of Agentic WordPress]] - **Author:** Noel Tock - **URL:** https://www.accelerateplugin.com/blog/the-5-levels-of-agentic-wordpress/#article - **Type:** article - **Citekey:** accelerate-5-levels-agentic-wp ## Links - [[3. Resources/MOCs/AI & the Future of Knowledge Work MOC.md]] - [[3. Resources/MOCs/AI Agent Safety MOC.md]] ## Author's thesis **Q1 (What is the author saying?):** Agentic WordPress work progresses through five distinct levels — from static-file coding assistance to a self-optimizing system that reacts to real-world outcomes — and the WordPress-native Abilities API plus MCP adapter is what lets agents act on live sites rather than just codebases. **Q2 (Why should the reader care?):** It reframes a widely-assumed cause-and-effect relationship with concrete, checkable evidence, which changes what action is worth taking next. ## Key arguments 1. The real inflection point (L2→L3) is a mentality shift from pointing agents at static files to letting them touch live staging/production sites. 2. The Abilities API functions as an agentic version of the REST API, with the MCP adapter as the highway it runs on; WP-CLI/SSH remains the fallback. 3. Level 4 is a sealed, automated world; Level 5 is the same setup dropped into the real world where outcomes push back and inform the site's own reaction. ## Claim atoms ### Agentic WordPress begins when agents touch live sites > [!quote] ID:rw_accelerate-5-levels-agentic-wp_1 > Many people limit themselves to working on a codebase, on so-called static files, but never point their coding agent at more dynamic tasks, such as interacting with your staging or production sites. - Section: Level 03 — Delegation. - Why this matters: names the real inflection point (L2→L3) as a mentality shift from static files to live sites. ### The Abilities API is the agentic REST API; MCP adapter is its highway > [!quote] ID:rw_accelerate-5-levels-agentic-wp_2 > The Abilities API is almost an agentic version of the REST API… The MCP adapter is the highway the Abilities API runs on. - Section: Level 03 — Delegation. - Why this matters: clearest one-line mental model of the WordPress-native agent stack. ### Level 4 is a sealed world; Level 5 lets outcomes push back > [!quote] ID:rw_accelerate-5-levels-agentic-wp_3 > The previous level, level four, is essentially the Truman Show: everything is alive, but it's a sealed world… Level five is that same setup dropped into the real world, where outcomes push back and inform your website how to react accordingly. - Section: Level 05 — Evolution. - Why this matters: sharpest distinction in the piece — automation vs. a self-optimizing system with a real feedback signal. ### Code is no longer the blocker; context and reliability are > [!quote] ID:rw_accelerate-5-levels-agentic-wp_4 > Code is not the blocker. The complexity comes later when you're trying to manage tokens, context, reduce errors, increase output quality. - Section: Summary — The way forward. - Why this matters: relocates the real difficulty from writing code to operating reliable loops. ## Permanent candidates #### Feedback Signals Make Automation Self-Improving An automated system only becomes self-improving once it is given an external outcome signal that can change its future decisions — without that feedback loop, it is merely scheduled, however lifelike it looks. ## Processing checklist - [x] Bib capsule complete - [x] Thesis captured (Q1 + Q2) - [x] Key arguments extracted (≤3) - [x] Permanent note candidate(s) spun out - [ ] Cross-links reviewed by next sweep