MCP WordPress Server

The MCP‑Wp server brings the Model Context Protocol to WordPress, allowing AI assistants such as Claude or Gemini to query and manipulate a site’s content through the familiar WordPress REST API. By exposing an MCP endpoint (), the plugin turns any WordPress installation into a first‑class AI data source, eliminating the need for custom connectors or manual API wrappers.

At its core, the server implements the Streamable HTTP transport defined in the latest MCP specification. This transport delivers data as a continuous stream, enabling large responses (e.g., full post archives or media libraries) to be consumed incrementally by the AI client. The implementation relies on the package, which handles the heavy lifting of protocol negotiation and request routing. As a result, developers can focus on defining WordPress‑specific resources—posts, pages, custom post types—and let the MCP layer manage the communication details.

Key capabilities include:

• Resource discovery: AI clients can list available WordPress entities (posts, pages, media) and their schemas.
• Tool execution: The server can expose custom tools such as “publish post” or “update page metadata,” allowing assistants to perform content operations directly.
• Prompt integration: Pre‑configured prompts tailored for WordPress tasks (e.g., “Generate a SEO‑friendly excerpt”) can be served to the AI, streamlining workflow.
• Sampling control: Clients can request specific data subsets or pagination, reducing bandwidth and improving response times.

Real‑world scenarios that benefit from this server are plentiful. A marketing team can let an AI assistant draft, edit, and publish blog posts on demand, while a developer can use the server to fetch structured data for generating documentation or analytics dashboards. In educational settings, students could interact with a WordPress site via conversational AI to learn about content management without exposing the raw REST API.

Integration is straightforward: once the plugin is activated, any MCP‑compatible client can register the site’s endpoint and start sending requests. The companion WP‑CLI AI command further simplifies this process, allowing developers to add the server to their local or production environment with a single command and immediately invoke AI actions from the terminal. This tight coupling between WordPress, MCP, and CLI tooling gives developers a powerful, low‑friction pathway to embed AI capabilities into their content workflows.