Drupal MCP. Control headless content and taxonomy via natural conversation.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Drupal MCP Server gives your AI client full control over headless Drupal content. You can list content nodes, manage taxonomy terms, provision new entities, and audit user accounts directly from natural conversation.
It handles complex JSON:API workflows, letting you treat Drupal like a standard data source, not a locked-down CMS interface.
What your AI agents can do
Create cms node
Writes a new content record into Drupal by providing a structured JSON payload.
Get file metadata
Retrieves specific structural details about a managed file.
Get single node
Fetches all content details for one specific node ID.
Use create_cms_node to send a structured JSON payload and write a new record into a Drupal entity, or use patch_cms_node to update a specific draft node.
List all content nodes using list_content_nodes, or fetch the full data for one node using get_single_node by its UUID.
Enumerate available categories using list_term_vocabularies, and then extract specific term properties with get_taxonomy_term.
Inspect managed files using list_managed_files, and verify the structure of a specific file using get_file_metadata.
List all registered users and their roles using list_drupal_users to audit the system's access control.
Irreversibly delete nodes using wipe_cms_node to clear live document entities.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
Drupal MCP Server: 10 Tools for Content & File Ops
These 10 tools let your agent perform every major content operation on your Drupal site, from creating nodes to auditing users, all through simple commands.
019d7589create cms node
Writes a new content record into Drupal by providing a structured JSON payload.
019d7589get file metadata
Retrieves specific structural details about a managed file.
019d7589get single node
Fetches all content details for one specific node ID.
019d7589get taxonomy term
Extracts the properties for a specific content category term.
019d7589list content nodes
Lists all content nodes available in the Drupal content structure.
019d7589list drupal users
Identifies a list of all active users and their access roles in the system.
019d7589list managed files
Inspects the internal records of all managed files in the site.
019d7589list term vocabularies
Lists all available content category groupings (taxonomies) on the site.
019d7589patch cms node
Safely updates a draft content node by replacing specific data segments.
019d7589wipe cms node
Permanently deletes a content node and its associated live data.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with Drupal, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
Drupal MCP Server - Manage Content with AI
Yo, this server hooks up your AI agent to your Drupal site's core data. You'll get full, programmatic control over all your content, bypassing the usual admin UI. It handles complex JSON:API workflows, so you can treat your Drupal data like any standard source, not some locked-down CMS. Your agent can read, write, and update content, manage user accounts, and audit files, all just by talking to it.
Content Management
To start, your agent can list every content node available using list_content_nodes, or it can grab all the specific details for just one node with get_single_node by its UUID. When you need to write new stuff, use create_cms_node to send a structured JSON payload and drop a new record into a Drupal entity.
If a draft node needs tweaking, patch_cms_node lets you safely update it by replacing specific data segments. You can also wipe a node entirely, permanently deleting it and all its associated live data using wipe_cms_node.
Taxonomy and Categories
Need to know what categories you've got? list_term_vocabularies lists all available content category groupings (taxonomies). Once you know the grouping, get_taxonomy_term pulls out the specific properties for a single category term.
Media and Files
To look at the files managed by the site, list_managed_files inspects the internal records of every file. If you need to check the structural details of a specific file, get_file_metadata retrieves that data.
User Accounts
Your agent can list every registered user and their roles in the system using list_drupal_users, letting you audit the site's access control.
This setup lets your agent handle everything from creating new content to clearing out old nodes, and it does it all through natural conversation.
How Drupal MCP Works
- 1 Subscribe to the server and provide your Drupal Base URL and authentication header. Make sure the JSON:API module is enabled on your site.
- 2 Give your AI client a natural language command (e.g., 'Find all articles about widgets').
- 3 The agent runs the necessary tools (e.g.,
list_content_nodes) and returns the structured data for you to act on.
The bottom line is: you manage complex Drupal content structures using plain language, without needing to write API calls.
Who Is Drupal MCP For?
This is for developers and content managers who spend too much time manually navigating CMS backends or writing boilerplate code just to read a simple list of articles. If you're constantly jumping between a developer IDE and a web admin panel, this saves you time.
Testing and debugging JSON:API endpoints or verifying entity schemas directly from your IDE. You use tools like get_single_node to validate data paths.
Creating and updating content nodes and managing taxonomy terms without ever logging into the Drupal admin UI.
Auditing registered users and monitoring managed file storage in real-time using list_drupal_users and list_managed_files.
What Changes When You Connect
- Writing content nodes used to mean logging into the CMS admin. Now, you can create or update content nodes using
create_cms_nodeorpatch_cms_nodesimply by talking to your agent. It bypasses the entire UI. - Checking a file's status or finding its CDN URI is a pain. Use
get_file_metadatato pull the exact structural match for any managed file, giving you the raw data you need immediately. - Need to see all the categories used on the site? Instead of clicking through every menu, run
list_term_vocabularies. It lists every structured rule defining your content's taxonomy instantly. - Auditing who has access is critical. Run
list_drupal_usersto get a precise array of all registered admins and editors, exposing their metadata and access roles instantly. - Debugging content structure is complex. Use
get_single_nodeto fetch the full details for a specific node UUID, eliminating the need to manually copy IDs and paste them into various endpoints. - Managing content lifecycle is safer now. You can permanently delete nodes using
wipe_cms_node, or safely adjust drafts withpatch_cms_node, keeping your live data intact until you're ready.
Real-World Use Cases
Refreshing an Article Draft
A content editor needs to update an article published weeks ago. Instead of finding the article in the CMS and clicking through multiple menus, they tell their agent: 'Update node UUID X with new text.' The agent runs patch_cms_node, safely replacing only the necessary text segment without touching other fields.
Schema Validation for New Content
A full-stack developer is integrating a new service. They don't trust the CMS UI. They run list_content_nodes to get a sample list, then use get_taxonomy_term to verify the exact structure of the required category terms, ensuring their payload matches the live schema.
Disabling Old Content
The Ops team needs to archive a large set of old, irrelevant articles. They run list_content_nodes to identify the target nodes, and then use wipe_cms_node to permanently delete them, keeping the system clean and audit-ready.
Checking Media Availability
A marketer needs to know if a specific image file is ready for the front end. They run list_managed_files to get the file ID, and then get_file_metadata confirms the raw CDN URI, solving the common headless media blocking problem.
The Tradeoffs
Calling the API for everything
Manually calling list_content_nodes repeatedly, then for each result, calling get_single_node to get the full data. This hits rate limits and is slow.
→
First, run list_content_nodes to get the list. If you need full data, pass the required UUIDs in a single, structured request to your agent, letting the agent orchestrate the retrieval.
Assuming content is always ready
Trying to publish content immediately after creation, only to fail because a related taxonomy term was missing or outdated.
→
Before creating, run list_term_vocabularies to check all available categories. Then, use create_cms_node with only the confirmed, existing term IDs.
Ignoring user permissions
Attempting to run a data synchronization job as a service account without checking if the account has write permissions for the specific content type.
→
Always check the user permissions first. Use list_drupal_users to verify the existence and roles of the service account before attempting any write operation like create_cms_node.
When It Fits, When It Doesn't
Use this server if you need to treat your Drupal site as a pure data source—a JSON:API endpoint—and your workflow involves complex content relationships (node ↔ file ↔ term). You need to programmatically read or modify content nodes, manage taxonomy, or audit users.
Don't use this if you only need to read basic, static site information (like a list of blog titles). For simple read operations, other generic data connectors might suffice. But if the data involves entity schemas, revisions, or complex file structures, this is the only toolset that gives you the depth needed.
It's crucial to remember that list_content_nodes gives a list of identifiers, while get_single_node gets the full content. Don't confuse the two.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Drupal. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Manually checking content structures is a total pain.
Right now, figuring out what data is available on a Drupal site means jumping through hoops: checking the admin UI for nodes, then switching to a file manager to see the media assets, and finally running a separate report to audit user access. It's clicking, copying IDs, and constantly context-switching.
With the Drupal MCP Server, you tell your agent exactly what you need. Want to see all content nodes and their categories? Just ask. The agent runs `list_content_nodes` and `list_term_vocabularies`, and you get the structured data back immediately. No more clicking through tabs.
Drupal MCP Server: Manage content and files via AI.
The biggest time-sink is the manual content lifecycle: draft changes, publishing, and file updates. You used to have to log in, find the node, click 'save draft,' manually update the field, and then click 'publish.'
Now, you just say, 'Update this article,' and the agent runs `patch_cms_node`. It handles the safe update, replacing only the necessary segment, and leaves everything else untouched. It's that simple.
Common Questions About Drupal MCP
How do I list all content nodes using the list_content_nodes tool? +
Run list_content_nodes and specify the required filters. This function identifies the boundaries of all content nodes in the headless Drupal model, giving you a list of available UUIDs to work with.
Can I update an article using the patch_cms_node tool? +
Yes. patch_cms_node allows you to modify a draft node by sending a structured payload that replaces only specific segments, ensuring you don't accidentally overwrite other data.
What if I need to delete a node permanently? Should I use wipe_cms_node? +
Yes, wipe_cms_node is the function for irreversible deletion. It permanently removes the node and its associated live document entities from the system.
How do I check the available content categories? Use list_term_vocabularies. +
Use list_term_vocabularies to enumerate all structured rules defining how content is categorized. This gives you the full list of taxonomies available on the site.
Is it possible to check file metadata with get_file_metadata? +
Yes. get_file_metadata pulls the exact structural matching for a specific file ID, which helps you find the raw CDN URI and verify the file's storage location.
How do I check which users have access to my Drupal site using the list_drupal_users tool? +
The list_drupal_users tool returns an array of all registered accounts. You'll see details like the user's name, email, and specific roles assigned, letting you audit who has access.
What data can I extract about content categorization using the get_taxonomy_term tool? +
The get_taxonomy_term tool performs structural extraction of properties for a specific term. It provides details about how that content is categorized, including its parent vocabularies.
Is there a way to find all managed files and their URIs using the list_managed_files tool? +
Yes, the list_managed_files tool pulls deep internal arrays of all configured files. This gives you raw CDN URIs, which is perfect for bypassing typical headless media restrictions.
Can my agent create a new content node in Drupal? +
Yes. Use the 'create_cms_node' tool. Provide the node type (e.g., 'article') and a JSON object with the attributes. The agent will orchestrate the HTTP POST request to your Drupal JSON:API to persist the entity.
How do I list all taxonomy terms for a specific vocabulary? +
Use the 'list_term_vocabularies' tool. Provide the vocabulary ID (e.g., 'tags'). Your agent will execute bulk iterations to track explicitly registered terms and return them within your chat context.
Can I retrieve the direct URL for a managed image file? +
Absolutely. Use the 'get_file_metadata' tool with the file ID. The agent will analyze the explicit UUID bounds and fetch the literal URL parameters, providing the public location of the managed attachment.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
SonarQube & SonarCloud
Bring your standalone or cloud SonarQube quality gates native to your AI logic. Find bugs, duplications, and rewrite vulnerable code instantly.
Comet ML
Manage machine learning experiments via Comet — track model metrics, audit project workspaces, and inspect ML run parameters directly from any AI agent.
Pulumi
Manage cloud infrastructure via Pulumi — list stacks, track deployments, audit outputs and tag resources from any AI agent.
You might also like
Chatwoot
Manage customer engagement via Chatwoot — track conversations, reply to messages, and manage contacts directly from any AI agent.
Upstream Lens
Monitor upstream oil and gas operations with production data analytics, well performance tracking, and field reporting tools.
BLS JOLTS — Job Openings, Quits & Turnover
Deep-dive into the 'Great Resignation' and labor tightness with the US Job Openings and Labor Turnover Survey (JOLTS). Track hires, fires, and voluntary quits.