From API to Agent: How WordPress 7.0 and Claude Code Are Rewriting Website Management

This revised guide preserves useful historical context while adding a dated maintenance review, current-source boundaries, and a practical fallback when a tool or platform has changed.

From API to Agent: How WordPress 7.0 and Claude Code Are Rewriting Website Management A WordPress site is no longer just a CMS — it’s an AI‑driven command center.

Think about the last time you logged into WordPress. You tweaked a meta description, resized a featured image, or fixed a broken internal link. Now imagine telling an AI to handle all of that while you stay in your terminal. That shift — from manual clicking to conversational command — is exactly what WordPress 7.0 and tools like Claude Code are bringing to the table.

This isn’t about “AI as a writing assistant” anymore. It’s about AI as an operator. WordPress 7.0 (codenamed “Armstrong”) built the plumbing. The ecosystem is now filling in the rest.

Here’s what changed, why it matters, and where to go if you want to start building your own AI‑driven WordPress workflow.

1. The New Foundation: WordPress 7.0’s AI Infrastructure

The Connectors API: One Key to Rule Them All

Before WordPress 7.0, adding AI capabilities meant juggling API keys across a dozen plugins. Each had its own settings page, each stored credentials differently. WordPress 7.0 introduced the Connectors API, a centralized credential management system for AI services. You can find it at Settings → Connectors in your wp‑admin.

The core change is simple: enter your OpenAI, Anthropic, or Google Gemini API key once. Every plugin that supports the new WP AI Client can use it — no reconfiguration needed. It’s standardized across the ecosystem.

📖 Official Field Guide WordPress 7.0 Field Guide — the complete developer reference for what’s changed.

🔌 Connectors API Deep Dive Introducing the Connectors API — auto‑discovery, authentication flows, and plugin developer notes.

🛠️ Hands‑On Guide What the AI Connectors Actually Do — a practical walkthrough of the new settings page.

The Abilities API: When AI Can Actually “Do” Things

Plugging in API keys is one thing. Letting AI operate your website is another. WordPress 7.0’s Abilities API gives AI agents structured access to core functionality: publishing posts, adjusting settings, navigating admin screens. The shift is subtle but profound — AI moves from “suggest” to “execute.”

🎥 WordPress.tv Talk: WordPress Development and Management with Claude Code — live demos of AI using the Abilities API, MCP, and Skills.

2. The Bridge: Connecting Claude Code to WordPress

The most practical way to start using AI agents with WordPress right now is through Application Passwords — WordPress’s native API authentication system. Generate one at Users → Profile → Application Passwords . Copy the token, and any AI agent (Claude Code, Cursor, custom scripts) can securely interact with your site via the REST API.

🔐 Step‑by‑Step Setup Publishing Content to WordPress with Claude Code Skills & the REST API — a 15‑minute guide: generating Application Passwords, creating a .env file, and building reusable Claude Code Skills.

🛡️ Security Best Practices Give Claude access to your WP website – but keep track of what they do — why Application Passwords are safer than sharing your main credentials, plus logging and auditing strategies.

3. The Advanced Frontier: Agent Skills and Custom Workflows

This is where things get genuinely interesting. WordPress officially released agent‑skills — portable bundles of instructions, checklists, and scripts that help AI assistants understand WordPress development best practices. The repository currently includes 13 core skills covering block development, theme.json, REST API usage, security patterns, and more.

If you work in VS Code with Claude Code, you’re in the most powerful configuration. You can define custom Skills in your project’s .claude/ directory, execute WP‑CLI commands directly, and build specialized agents for SEO auditing, image generation, internal linking, or content governance — all without leaving your terminal.

📦 Official GitHub Repository WordPress agent‑skills on GitHub — the definitive source, updated with 1,200+ stars and active community contributions.

🧠 Case Study — 6 Agents, 16,000 Posts I Built a WordPress Plugin with a Team of 6 AI Agents — processing 16,000 posts in 90 seconds using Claude Code’s Agent Teams.

📈 SEO Automation Case Study Automating WordPress SEO Optimisation with MCP and Claude Code — fixing 11 broken URLs, adding meta descriptions to 52 pages, and implementing strategic internal linking across 40 posts, all without opening the WordPress admin interface.

What This Means for You

This shift — from “AI suggests, you click” to “you command, AI executes” — represents a fundamentally different way of managing websites. The plugins you rely on will increasingly need to support the new WP AI Client and Connectors API to remain competitive. But more importantly, you now have the tools to build your own agents: custom workflows that scan, optimize, link, and publish without manual intervention.

Whether you start with Application Passwords and a simple Claude Code Skill, or dive into the full agent‑skills repository, the path is open. The question is no longer if AI can run parts of your site — it’s what parts you’re ready to hand over .

🚀 Your Learning Path 1. Start small — Set up an Application Password and try the REST API publishing workflow from the Anicca guide. 2. Explore the official resources — Read the Connectors API documentation and watch the WordPress.tv Claude Code talk. 3. Go deep — Clone the agent‑skills repository, study the case studies, and build your first custom SEO audit Skill.

📝 This post was inspired by a detailed internal analysis of WordPress 7.0’s AI infrastructure and the growing ecosystem of agent‑based WordPress tools. All linked resources were verified as of May 2026.

flowchart LR subgraph User[“👤 User”] CLI[“Terminal / Claude Code”] Skills[“Custom Skills”] end subgraph WP[“🏠 WordPress 7.0”] REST[“REST API”] Auth[“App Passwords”] Connectors[“Connectors API\n(AI keys)”] Abilities[“Abilities API\n(Execute)”] end subgraph AI[“🤖 AI Agents & Services”] Claude[“Claude Code”] OpenAI[“OpenAI”] Anthropic[“Anthropic”] Gemini[“Google Gemini”] end subgraph Tasks[“📋 Workflows”] SEO[“SEO Fixes”] Content[“Publish/Edit”] Linking[“Linking”] Meta[“Meta Desc”] end CLI –>|App Password| Auth Auth –> REST REST –> Connectors & Abilities Connectors –> OpenAI & Anthropic & Gemini Abilities –> Tasks CLI –> Skills –> Claude Claude –>|agent-skills & REST| Tasks style WP fill:#5b21b6,color:#fff style User fill:#f97316,color:#000 style AI fill:#c084fc,color:#000 style Tasks fill:#10b981,color:#000

📚 Related Guides

Review status — last checked August 3, 2026

  • Originally published or scheduled: 2026-05-21
  • Last reviewed: August 3, 2026
  • Maintenance status: current / version-specific
  • Editorial action: WordPress guidance must name the version and be rechecked against the current release archive, developer notes, plugin requirements, and the site’s actual installed version.

This date is a maintenance signal, not a promise that the product, platform, price, model, or policy will remain unchanged. When a reader is deciding whether to adopt a tool, the current official documentation, account experience, terms, pricing page, and support channel take priority over this article.

What changed since the older version?

Older AI-content articles often combine three different kinds of information: a durable workflow idea, a product-specific observation, and a time-sensitive claim. The workflow may still be useful while the interface, model, limits, price, integration, or policy has changed. This revision separates those layers. Use the durable part as a method, use the dated status above to understand the review boundary, and re-test every product-specific step before using it in production. A screenshot from 2022 is evidence of what was visible then; it is not evidence of what a new account sees today.

How to use the method safely

Start with the outcome, audience, source material, and acceptance criteria. Keep a record of the prompt, input facts, model or tool name, date, output, human edits, citations, and final approval. Check claims against primary sources. Do not publish generated text just because it is fluent. Review factual accuracy, attribution, privacy, copyright, brand voice, search intent, disclosure, and whether the result could mislead a reader about a product or service. If a tool is unavailable, replace it with a documented workflow rather than pretending that an old button or endpoint still exists.

What should be rechecked before publication?

  • Open the official product, documentation, changelog, pricing, and support pages.
  • Confirm the current sign-in path, plan name, feature name, API endpoint, plugin version, or policy section.
  • Run the smallest reproducible test and record the date, account type, region, and result.
  • Remove unsupported claims, stale screenshots, dead links, and prices without a current source.
  • Add a visible last-reviewed date and label the article as current, changed, unverified, or retired.
  • Give readers a fallback path when the tool is discontinued, inaccessible, or unsuitable.

If the tool is no longer maintained

Do not silently rewrite history. Keep the original use case clear, label the tool as legacy or unverified, remove conversion language that implies current availability, and explain what a reader should use to evaluate a replacement. A replacement is not automatically equivalent: compare inputs, outputs, export formats, privacy terms, integrations, rate limits, support, and migration effort. If no trustworthy replacement has been verified, say so. That is more useful than filling the gap with an invented recommendation.

Current source boundary

This site publishes practical editorial guidance, not a guarantee of a vendor’s uptime, pricing, performance, or future roadmap. Tool status is checked on the date above and should be reviewed again before a purchase, migration, API deployment, or compliance decision.

Sources checked for this review

Related AI content guides

Leave a Reply

Your email address will not be published. Required fields are marked *

Important updates waiting for you!
Consectetur eget cras neque augue malesuada urna urna hendrerit tellus.