Directus MCP. Manage your database content conversationally.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
Directus MCP connects your AI client straight to your SQL database, turning complex content models into natural conversation. Instead of writing API calls or navigating a dashboard, you talk to it and get data back instantly.
Check collection schemas, create new articles, manage user roles, and run audit reports—all through your agent. It's direct, conversational control over structured data.
What your AI agents can do
Check directus status
Verifies that the connection to your Directus database is working correctly.
Create item
Creates a new record or item in any specified content collection.
Delete item
Permanently deletes an existing record from a collection.
Fetch specific records, list all items in a collection, or search across content by keyword.
View which data collections exist and inspect the specific fields (the schema) for any chosen model.
Create new content entries or update existing ones by providing structured JSON payloads.
Browse uploaded media assets, get file lists, and retrieve the metadata for any specific file.
List all registered users, view their assigned roles, or check your own profile status.
Access the audit trail to see a log of recent database operations for compliance checks.
Ask AI about this MCP
Supported MCP Clients
OAuth 2.0 Compatible Gemini Waiting for input…
Directus: Data Management Tools with 16 Tools
These tools let you perform every core database operation—from listing users to creating new articles—all through your AI agent.
Make your AI actually useful.
Add this MCP to Claude, Cursor, or Windsurf and your AI stops guessing. It gets real tools to look things up, take action, and handle the stuff you keep doing by hand.
Start using Directus on Vinkius019dd0e0check directus status
Verifies that the connection to your Directus database is working correctly.
019dd0e0create item
Creates a new record or item in any specified content collection.
019dd0e0delete item
Permanently deletes an existing record from a collection.
019dd0e0get collection
Retrieves detailed information about a specific data collection model.
019dd0e0get file
Gets the metadata and details for an uploaded media asset.
019dd0e0get item
Retrieves all available field data for a single record by its ID.
019dd0e0get me
Returns the profile details and information for the currently authenticated user.
019dd0e0list activity
Lists recent operations across the database, acting as an activity log.
019dd0e0list collections
Returns a list of all available data collections in your system.
019dd0e0list fields
Lists every field and its data type for a specific, chosen collection.
019dd0e0list files
Retrieves a list of all uploaded media files in the system.
019dd0e0list items
Fetches a paginated list of items from a designated content collection.
019dd0e0list roles
Lists all user roles defined in the system (e.g., Admin, Editor).
019dd0e0list users
Retrieves a list of every user account and their current status.
019dd0e0search items
Searches for content items across collections using specific keywords or criteria.
019dd0e0update item
Modifies the data fields of an existing record by its ID.
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 Directus, then connect any of our 4,900+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,900+ 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
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Directus. 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 16 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
The Dashboard Click Tax
Today, if you need to check content and then make a change, you're stuck in an operational loop. You open the Directus dashboard, navigate to 'Articles,' filter by status, copy the article ID, switch tabs to 'Users' to see who can edit it, and finally return to the Articles tab to manually update the record. It’s tedious clicking, copying, and switching context.
With this MCP, you talk to your agent. You tell it what needs doing—like 'Find all articles assigned to Sarah that need a status change.' Your agent handles the multi-step process: listing users, checking roles, fetching items, and executing `update_item`. The result is immediate and conversationally presented.
Directus MCP Gives You Conversational Control
You eliminate manual data lookups. Instead of opening a file management view to see if an asset exists, you simply ask the agent, 'What media files were attached to article X?' The tool handles the `get_file` lookup and returns the details instantly.
Your process becomes linear: one prompt, multiple backend operations. You don't just read data; you govern it. It’s direct control over your content lifecycle.
What you can do with this MCP connector
Manage the core content structure of your application without touching a dashboard or writing boilerplate code. This MCP lets you treat your database like a conversation partner. You ask your AI client for something specific—say, 'Show me all articles published last month that mention AI'—and it handles the necessary reads and filters.
It’s pure data interaction via natural language. When you connect this through Vinkius, your agent gets access to thousands of other tools, making it a central hub for managing content from any MCP-compatible client. You can inspect what fields exist in a collection using list_fields or even build out new records by calling create_item.
Whether you're a developer debugging data issues or a content manager pushing out an update, this connection gives you immediate control over your entire content ecosystem.
019dd0e0-fcc8-708e-8646-61ae1562985e 
How Directus MCP Works
- 1 Subscribe to this MCP and enter your Directus Static Token into the Vinkius catalog.
- 2 Connect your AI client (Claude, Cursor, etc.) to the MCP. The agent now knows how to talk to your database.
- 3 Ask a natural language question: 'What are the fields in my articles collection?' or 'Create an item with title X.' You get instant data results.
The bottom line is, you use your AI client's chat interface to manage complex backend operations without needing to open a single administrative dashboard.
Who Is Directus MCP For?
This MCP targets the content team who needs to publish articles but doesn't want to log into Directus. It’s perfect for developers debugging data models or system administrators checking user permissions without running manual SQL queries.
Needs to create, read, and update articles and pages quickly by telling the agent what content needs publishing.
Requires inspecting collection schemas using list_fields or debugging data integrity by reviewing the full audit trail via list_activity.
Needs to manage user access, check who has what roles with list_users, and verify system health using check_directus_status.
What Changes When You Connect
- Instead of writing raw API calls, you can use your agent to run
list_collectionsand instantly see every data model available in the system. It's faster than navigating the dashboard. - Need to publish an article? You don't have to open Directus; simply instruct your agent to
create_item, supplying all necessary fields, like title and body content, through conversation. - Debugging a complex data issue is easier when you can call
list_fieldsto inspect the exact schema of any collection. This prevents guessing what data types are available. - When reviewing compliance, instead of running manual reports, ask your agent to use
list_activityto summarize all changes made in the last few hours. - You gain full control over user accounts; you can list who needs access via
list_usersand then manage their roles using other tools within Vinkius.
Real-World Use Cases
Content writer blocked on a new article draft.
The writer asks the agent, 'Find all articles about finance that are currently published.' The agent runs search_items and provides a list of three links. They then use get_item to pull the raw text body from one of those drafts for reference.
Developer needs to add a new content type.
The developer runs 'What collections do I have?' (using list_collections). Seeing the missing 'Podcast' collection, they then use get_collection on an existing model like 'articles' to understand how field types are structured before creating the new one.
Admin needs to onboard a new team member.
The admin asks the agent to check user roles. The agent calls list_users and identifies that the new hire only has 'Reader' status, so the admin uses conversational instructions to update their role.
System audit after a data breach scare.
The auditor asks for all database changes in the last 24 hours. The agent calls list_activity and summarizes every operation, providing an immediate compliance log without leaving the chat interface.
The Tradeoffs
Treating it like a single form.
A user thinks they can just send raw data to update_item without knowing the required field IDs or status codes, leading to 400-level errors.
→
Always start by calling get_collection and then list_fields. This confirms the exact structure you need before attempting any write operation. If you're unsure of a user ID, use list_users first.
Running random CRUD operations.
A developer tries to delete an item using delete_item without verifying if the content is already referenced by another collection. This risks breaking data dependencies and losing critical info.
→
Before deleting, always run a search or list command first—for instance, use search_items on the parent collection. Verify that no other related items depend on the record you want to delete.
Assuming field existence.
A content manager tries to write a new article and uses a custom field name (e.g., 'Marketing Focus') that isn't defined in the schema, causing the create_item call to fail.
→
Use list_fields on the target collection first. This shows you the precise names of every available field, ensuring your agent payload uses the correct keys.
When It Fits, When It Doesn't
You need this MCP if managing content requires multiple steps: reading metadata, checking user access, and then writing new records based on that data. It's perfect for complex workflows like 'Find all drafts -> Get their author ID -> Update the status to published.' Don't use it if your only task is a simple read (like just viewing one item). For basic reads, direct API calls are fine. But when you need to orchestrate multiple actions—for example, listing users and then using that list to update roles—this conversational approach through Vinkius is the right choice. It wraps those complex interactions into simple chat commands.
Common Questions About Directus MCP
How do I use `create_item` with the Directus MCP? +
You provide the required fields and values in natural language, letting the agent structure the JSON payload for you. You don't need to know the exact API body; just tell it what data needs creating.
What is the difference between `list_items` and `search_items`? +
list_items gives you a paginated list of content from an entire collection. Use search_items when you know specific keywords or criteria, allowing for much finer filtering.
Can I check the connection using `check_directus_status`? +
Yes, calling check_directus_status simply confirms that your AI client has a live and working connection to your Directus instance. It's a quick health check.
How does `list_users` help me manage content? +
list_users helps by showing who can access the data, which is critical when you need to know if an account has permission to create or modify a specific type of record.
How do I use `list_fields` to map out a collection's full data model? +
It immediately lists every field and its data type within a specified collection. This lets you inspect the entire schema, which is essential before writing any update or creation payload.
If I run `delete_item`, what steps should I take to avoid accidentally deleting critical content? +
The tool requires specific IDs and collection names for targeted deletion. Always verify the item details first using get_item so you know exactly what data you're about to remove.
What is the best way to browse and retrieve metadata for uploaded assets using `list_files`? +
It provides a comprehensive list of all stored files. You get details like file names, sizes, and metadata without needing to interact with the visual dashboard.
Does `list_activity` help me track compliance or debug recent database issues? +
Yes, it generates an activity log of operations that occurred recently. You can see who did what, when they logged in, and which collections were modified.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.