Ghost MCP Server – Overview

The Ghost MCP Server bridges the gap between AI assistants and a Ghost CMS instance by exposing the full breadth of the Ghost Admin API through the Model Context Protocol. It solves a common pain point for developers who want to let language models manage content, pages, members, and media without writing custom adapters or handling authentication themselves. By running a lightweight Node.js service that listens for MCP requests, an AI assistant can perform CRUD operations on posts and pages, query members, manage tags, and upload images—all in a single, well‑defined API surface.

At its core, the server translates standard MCP tool calls into Ghost Admin REST requests. For example, a call with title and HTML content is forwarded to the Ghost endpoint that creates a new blog entry. The server also handles pagination, search, and filtering for posts, pages, and members, providing rich query capabilities that would otherwise require manual construction of API URLs. This tight integration means developers can focus on higher‑level business logic while the MCP server manages authentication, rate limiting, and response formatting.

Key capabilities include:

• Post & Page Management – Create, read, update, delete, and search both posts and static pages with support for all Ghost content formats (HTML, Mobiledoc, Lexical).
• Member & Tag Handling – Retrieve lists of members or tags, perform searches, and include related data such as labels or newsletters.
• Image Upload – Accept base64‑encoded images and upload them to Ghost with optional purpose metadata, simplifying media handling.
• Author & Tag Management – Expose endpoints for author and tag CRUD operations, enabling comprehensive content governance.

Real‑world use cases are abundant. Content‑heavy teams can let an AI assistant draft articles, schedule publications, and tag them appropriately—all while keeping the workflow within Ghost. Marketing workflows can auto‑populate landing pages or member newsletters based on AI recommendations. Even simple automation scripts can use the MCP server to pull the latest posts for a newsletter or purge outdated content, all triggered from an AI prompt.

Integration with existing AI workflows is straightforward. Once the server is running, any MCP‑compliant client (Claude, Gemini, or custom tooling) can declare the available tools (, , , etc.) and invoke them as part of a conversational chain. The server’s stateless design ensures that each request is independent, making it ideal for cloud‑native deployments or local development environments. Its unique advantage lies in the seamless coupling of Ghost’s powerful CMS features with an AI assistant’s natural language interface, eliminating the need for bespoke SDKs or manual API calls.