- VALIDATION
- MAPPING
- RELIABILITY
- HUMAN
Short answer
Can my AI see my Directus data model?
Yes. Your AI agent uses this MCP to pull collection names and field definitions directly from your Directus instance. This lets you validate field types and relationships while you write code. You'll get a clear view of your data structure.
The results of inspection.
Where your data model lands.
The AI handles the heavy lifting of mapping your data model.
VALIDATION
Verified field types
The AI checks if a field is a string, integer, or relation. You'll know exactly what data types to expect in your queries.
MAPPING
Accurate collection lists
Your agent pulls the full list of available collections. This prevents typos and ensures you're targeting the right tables.
RELIABILITY
Real-time model sync
Since the AI reads the live schema, it always knows if you've added a new field or changed a relationship.
HUMAN
Manual schema design
The AI can read the structure, but you still need to decide how to build and modify the collections in the Directus dashboard.
The workflow
What your AI does when you ask about your data.
The agent follows a specific path to map out your Directus environment.
Find collections
The agent identifies every collection currently available in your Directus instance.
list_collectionsTarget a collection
Once you name a specific collection, the agent pulls its full metadata.
get_collectionInspect field details
The agent looks at the individual fields within a collection to see their types and properties.
list_fieldsVerify field types
The agent uses field information to confirm if a field matches your expected data format.
list_fields
Try it
Copy these to start.
Use these prompts to begin interacting with your Directus schema.
Starting points
Swap out the collection names for the ones actually used in your Directus project.
Directus Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_GfUHShsR6dDBscbzwfGCgCgJZgBT5MsQqEXOFrFu/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Directus capabilities are ready to use.
{
"mcpServers": {
"directus-alternative-mcp": {
"url": "https://edge.vinkius.com/vk_preview_GfUHShsR6dDBscbzwfGCgCgJZgBT5MsQqEXOFrFu/mcp"
}
}
}List all the collections in my Directus instance.
What are the fields inside the 'orders' collection?
Check the 'customers' collection and tell me which fields are required.
Does the 'products' collection have a foreign key relationship to 'categories'?
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Directus once, then ask.
Follow the link to connect your instance. Your credentials stay encrypted, and you'll be ready to chat with your data immediately.
Connect Directus to your AIFAQ
How the schema inspection works
- 01
Can the AI change my database schema?
No. This MCP only provides tools to read and list existing structures. It cannot create, update, or delete collections or fields.
- 02
Can the AI see the actual data inside my rows?
No. This specific task is limited to inspecting the schema and field definitions, not querying the records themselves.
- 03
Does the AI know about my custom fields?
Yes. If you added a custom field in the Directus dashboard, the agent will see it when it inspects the collection.
- 04
How does the AI know which fields are relations?
The agent reads the field metadata to identify relationship types, allowing it to understand how your collections link together.
- 05
Do I need to give the AI my Directus password?
No. You connect via Vinkius, which manages the connection securely. Your agent never sees your raw login credentials.
More questions about Directus? See everything the Directus Connector can do
Connect Directus to Claude, Cursor, ChatGPT & more
