Statamic MCP for AI. Access every piece of your site's structured data from chat.

You connect your AI client to this Statamic MCP Server and you get full access to every piece of data on your site's backend—no logging into the admin dashboard required. Your agent reads all your flat-file content and database settings using plain language prompts.

When you need to understand how your site is built, start by mapping the structure. Use get_nav_tree when you want the full, nested map of your main navigation menu. For content sections, run get_collection_tree to see the entire hierarchy of structured entries in a collection.

To pull article data, you've got options. You can use list_entries to grab paginated lists of articles from any defined collection; this tool supports filtering by field value or entry status, so you don't get junk data. If you just need everything for one specific piece, run get_entry.

For media management, your agent handles it too. Run list_assets to see every file inside a specified asset container or folder. To pull all the specifics on a single image or document, use get_asset with its unique ID.

When you're dealing with site-wide configuration, you need global data. First, run list_globals to get names of every Global set used across your site. Then, if you know which one you want, use get_global to fetch all the variables and settings stored within that specific set.

Want to check out how users are set up? Use list_users to pull a roster of every account configured in Statamic. To get detailed profile info or roles for just one person, run get_user.

For forms, you can't miss them either. Run list_forms to return a list of every submission form across the site. If you need to inspect the specific fields and structure of one particular form, use get_form.

Taxonomy and categorization are handled by two tools. Start with list_terms if you want to see all the available categories (or tags) within a specific taxonomy. When you know the exact term's details—like its description or usage count—run get_term.

It wraps up by giving you access to system-level lists: use list_forms for every form, and list_users for all accounts. Basically, it lets your AI client talk directly to the CMS API so you can pull everything—from individual article fields using get_entry, to checking which specific taxonomy term details with get_term, or listing multiple items from a collection using list_entries.

It's clean data, straight conversationally.

To summarize the process: You ask your agent for something—say, 'Give me all active blog posts in 2023 that mention X.' Your client runs the necessary API calls behind the scenes using list_entries, applies the filters and sorting you asked for, and spits out clean JSON data. It’s about direct access to structure: get_collection_tree maps the content flow; get_nav_tree maps the site's actual menus; and get_global pulls all the variable settings that drive the whole thing.**

Use your AI client to:

• List articles from any collection, supporting advanced filtering and pagination via list_entries.
• Inspect media assets by listing files (list_assets) or retrieving specific data points using get_asset.
• Map the site's entire structure: use get_collection_tree for content sections and get_nav_tree for main menus.
• Check configuration variables across the whole platform with list_globals and get_global.
• Query categories: list all available terms using list_terms, or get deep details on one term using get_term.

Your agent handles these systems calls so you don't have to touch the admin dashboard. You just talk to it.