WordPress launches the canonical MCP Adapter plugin, enabling AI agents to securely interact with websites via Anthropic's Model Context Protocol standard.

The bridge between static content management and autonomous AI agency has just become official. WordPress has released the canonical MCP Adapter plugin, providing developers and business owners with a standardized method to connect artificial intelligence systems directly to their websites. While the project had previously existed on GitHub—amassing over 40,000 installations—version 0.7.0 marks its formal debut in the official WordPress.org plugin repository.

Understanding the Model Context Protocol

At the heart of this release is the Model Context Protocol (MCP), an open standard developed by Anthropic. Rather than requiring a unique, proprietary integration for every different AI model, MCP provides a universal language that allows AI systems to securely connect to applications, data sources, and tools. For a WordPress site, this means an AI agent is no longer limited to simply reading a page via a crawler; it can now interact with the site's internal functions.

The adapter allows AI to engage with WordPress functions, such as PHP code snippets, to perform specific tasks like publishing posts or modifying site settings. By acting as a translator, the MCP Adapter enables AI clients—such as Claude Desktop or Cursor—to discover what a specific WordPress site is capable of and execute those actions within the permissions granted by the administrator.

The role of the Abilities API

To ensure that AI integration does not lead to chaotic or redundant operations, the MCP Adapter utilizes the Abilities API. This API, which has been part of the WordPress core since version 6.9, serves as a standardized list of actions a website knows how to perform.

Jason Adams, Director of Engineering, AI at Automattic, noted that the use of the Abilities API is critical for handling authorization and avoiding functional redundancy. Essentially, the AI agent asks the site what it can do, and the Abilities API provides the menu of available actions. This architecture ensures that the AI does not attempt to execute commands that the site cannot support or that the user has not authorized, maintaining a layer of security between the LLM and the server.

Prioritizing backward compatibility

One of the most significant technical decisions in the development of the MCP Adapter is the focus on backward compatibility. In the fast-moving world of AI, protocols evolve weekly, but WordPress plugins are often expected to remain functional for years.

The backwards compatibility matters. WordPress plugins tend to stay around for years.

Adams emphasized that while the WordPress team cannot control the evolution of the MCP protocol itself, they can control how their ecosystem handles those changes. By architecting the plugin for backward compatibility, WordPress ensures that businesses will not find their AI integrations broken every time a new version of an AI model or protocol is released.

A crowded market of AI connectors

The release of an official plugin does not mean the market is now a monopoly. The WordPress ecosystem already hosts various third-party solutions designed to bridge the gap between AI and web management. For instance, SeedProd developed WPVibe (also known as Vibe AI), an MCP server that allows users to connect any MCP-ready AI to their site.

WPVibe focuses heavily on the user experience for non-developers, offering features like a sandbox environment where users can edit theme files and preview changes before pushing them to a production website. It employs AES-256-GCM encryption for credentials and relays tool calls via the WordPress REST API, positioning the AI as a co-pilot for SEO and site redesign.

Verification of official status

For entrepreneurs and developers, distinguishing between a community plugin and an official tool is vital for security and long-term support. The MCP Adapter is confirmed as official through several markers: its authorship is listed as WordPress.org, the source code resides in the official WordPress GitHub repository, and it originated from the AI Building Blocks for WordPress initiative.

Furthermore, Automattic has aligned its efforts with this canonical version, archiving its parallel project (wordpress-mcp) and directing users toward the MCP Adapter and the official documentation at developer.wordpress.com. This consolidation signals a unified strategy for how the world's most popular CMS will handle the agentic web.

Global business implications for USA and UK markets

For businesses operating in the USA and UK, the shift toward MCP-enabled websites represents a transition from AI as a content generator to AI as an operational employee. The ability for an AI agent to autonomously manage a WordPress site—updating SEO metadata, publishing scheduled content, or adjusting layout based on performance data—reduces the operational overhead for lean entrepreneurial teams.

From a regulatory perspective, US and UK firms must focus on the authorization layer provided by the Abilities API. Because MCP allows AI to perform actions (not just read data), the risk of unauthorized site changes or data leaks increases. Companies should implement strict permission protocols to ensure that AI agents have the minimum necessary access to perform their tasks.

While the EU AI Act introduces stringent transparency and risk-management requirements for AI systems, US and UK businesses currently operate in a more flexible environment. However, the use of encrypted credentials and standardized protocols like MCP is a prudent step toward future-proofing sites against potential upcoming regulations regarding AI autonomy and data provenance.

FAQ

---

Sources: Searchenginejournal (2), Semasocial ·