Skip to main content
Glama

@staminna/directus-mcp-server

MCP server for Directus 12 — items, collections, files, flows, users, and schema tools. TypeScript, typed throughout.

npm version License: MIT CI mcp-server-claude MCP server

mcp-server-claude MCP server

Test Coverage

Statements

Branches

Functions

Lines

Statements

Branches

Functions

Lines

Coverage badges are generated from coverage/coverage-summary.json by npm run badges (no external service required). Run npm run test:coverage first.

Features

  • 🔐 Full Authentication - Token-based authentication with Directus

  • 📦 Collection Management - CRUD operations for collections and items

  • 📁 File Operations - Upload, download, and manage files

  • 🔄 Flow Management - Create, update, trigger, and manage Directus Flows

  • 👥 User Management - User CRUD and role management

  • 🔍 Schema Tools - Analyze and validate collection schemas

  • 🩺 Diagnostics - Collection access diagnostics and troubleshooting

Related MCP server: Storyblok MCP Server

Installation

npm install -g @staminna/directus-mcp-server

From Source

git clone https://github.com/staminna/mcp-server-claude.git
cd mcp-server-claude
npm install
npm run build

Environment Variables

Variable

Required

Description

DIRECTUS_URL

Yes

Your Directus instance URL (e.g., http://localhost:8065)

DIRECTUS_TOKEN

Yes

Static API token with appropriate permissions

DIRECTUS_PROMPTS_COLLECTION_ENABLED

No

Enable AI prompts collection (true/false)

DIRECTUS_PROMPTS_COLLECTION

No

Collection name for AI prompts (default: ai_prompts)

DIRECTUS_RESOURCES_ENABLED

No

Enable resources feature (true/false)

DIRECTUS_RESOURCES_EXCLUDE_SYSTEM

No

Exclude system collections from resources (true/false)

NODE_ENV

No

Environment mode (development/production)

DIRECTUS_TIMEOUT

No

Request timeout in ms (default: 30000)

DIRECTUS_RETRIES

No

Retry attempts for network errors, 5xx and 429 (default: 3)

DIRECTUS_RETRY_DELAY

No

Base backoff delay in ms (default: 1000)

DIRECTUS_MAX_RETRY_DELAY

No

Backoff ceiling in ms (default: 10000)

DIRECTUS_IMPORT_MAX_FILE_SIZE

No

Client-side import size ceiling in bytes, mirroring the Directus IMPORT_MAX_FILE_SIZE (default: 50 MB)

LOG_LEVEL

No

DEBUG/INFO/WARN/ERROR (default: INFO). Logs go to stderr; stdout is reserved for MCP

TLS / client certificates

Set these when the Directus instance uses a private CA or requires a client certificate. Each of CA/CERT/KEY/PFX accepts either a file path or the PEM/DER content itself.

Variable

Description

DIRECTUS_HTTPS_CA

Certificate authority

DIRECTUS_HTTPS_CERT

Client certificate

DIRECTUS_HTTPS_KEY

Client private key

DIRECTUS_HTTPS_PFX

PKCS#12 bundle (alternative to cert/key)

DIRECTUS_HTTPS_PASSPHRASE

Passphrase for the key or PFX

DIRECTUS_HTTPS_REJECT_UNAUTHORIZED

false to accept self-signed certificates

DIRECTUS_HTTPS_SERVERNAME

SNI server name override


Authentication — no OAuth required

This server uses a static Directus access token (DIRECTUS_TOKEN) and runs over stdio transport. OAuth is not required, by design:

  • The MCP specification only defines OAuth 2.1 authorization for HTTP-based transports. For stdio servers the spec says implementations "SHOULD NOT" use it and should instead retrieve credentials from the environment — exactly what this server does.

  • Directus 12 fully supports static access tokens. The OAuth 2.1 support Directus added (mid-2026) applies to its own built-in remote MCP endpoint and is optional; there are no breaking changes to token authentication in Directus 12 (see DIRECTUS_V12_BREAKING_CHANGES.md).

  • OAuth only becomes relevant if you expose an MCP server remotely over HTTP (Streamable HTTP/SSE). As a local stdio subprocess of Claude Desktop, Claude Code, Cursor, etc., this server needs only the env token.

Generate the token in Directus under User Settings → Token (use a dedicated user with least-privilege role for production).

Using with a Claude subscription (Max/Pro) — no API key needed

MCP servers do not consume Anthropic API tokens themselves; only the AI client's model calls do. If you use this server inside Claude Code or Claude Desktop with a Claude Max (or Pro) subscription, the model usage is covered by the subscription — you do not need an Anthropic API key. An API key is only required when driving Claude programmatically via the Claude API (e.g. the remote MCP connector).


IDE Configuration

🟣 Cursor

  1. Open Cursor Settings: Cmd+, (macOS) or Ctrl+, (Windows/Linux)

  2. Search for "MCP" or navigate to Features → MCP Servers

  3. Click "Edit in settings.json"

  4. Add the following configuration:

{
  "mcpServers": {
    "directus": {
      "command": "npx",
      "args": [
        "-y",
        "@staminna/directus-mcp-server"
      ],
      "env": {
        "DIRECTUS_URL": "http://localhost:8065",
        "DIRECTUS_TOKEN": "your-directus-token-here"
      }
    }
  }
}

Or if installed locally:

{
  "mcpServers": {
    "directus": {
      "command": "node",
      "args": [
        "/path/to/mcp-server-claude/dist/index.js"
      ],
      "env": {
        "DIRECTUS_URL": "http://localhost:8065",
        "DIRECTUS_TOKEN": "your-directus-token-here"
      }
    }
  }
}
  1. Save the file and restart Cursor


🌊 Windsurf

  1. Open Windsurf Settings: Cmd+, (macOS) or Ctrl+, (Windows/Linux)

  2. Search for "MCP Servers"

  3. Click "Edit in settings.json"

  4. Add the following configuration:

{
  "mcpServers": {
    "directus": {
      "command": "npx",
      "args": [
        "-y",
        "@staminna/directus-mcp-server"
      ],
      "env": {
        "DIRECTUS_URL": "http://localhost:8065",
        "DIRECTUS_TOKEN": "your-directus-token-here",
        "DIRECTUS_PROMPTS_COLLECTION_ENABLED": "true",
        "DIRECTUS_PROMPTS_COLLECTION": "ai_prompts",
        "DIRECTUS_RESOURCES_ENABLED": "true",
        "DIRECTUS_RESOURCES_EXCLUDE_SYSTEM": "true",
        "NODE_ENV": "production"
      }
    }
  }
}

Or if installed locally:

{
  "mcpServers": {
    "directus": {
      "command": "node",
      "args": [
        "/path/to/mcp-server-claude/dist/index.js"
      ],
      "env": {
        "DIRECTUS_URL": "http://localhost:8065",
        "DIRECTUS_TOKEN": "your-directus-token-here"
      }
    }
  }
}
  1. Save the file

  2. Quit Windsurf completely (Cmd+Q or Ctrl+Q)

  3. Reopen Windsurf and wait ~10 seconds for MCP to initialize


🤖 Claude Desktop

  1. Locate your Claude Desktop config file:

    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

    • Windows: %APPDATA%\Claude\claude_desktop_config.json

    • Linux: ~/.config/Claude/claude_desktop_config.json

  2. Create or edit the config file:

{
  "mcpServers": {
    "directus": {
      "command": "npx",
      "args": [
        "-y",
        "@staminna/directus-mcp-server"
      ],
      "env": {
        "DIRECTUS_URL": "http://localhost:8065",
        "DIRECTUS_TOKEN": "your-directus-token-here"
      }
    }
  }
}

Or if installed locally:

{
  "mcpServers": {
    "directus": {
      "command": "node",
      "args": [
        "/path/to/mcp-server-claude/dist/index.js"
      ],
      "env": {
        "DIRECTUS_URL": "http://localhost:8065",
        "DIRECTUS_TOKEN": "your-directus-token-here"
      }
    }
  }
}
  1. Save the file and restart Claude Desktop


🔮 Claude.ai (Web with MCP)

For Claude.ai web interface with MCP support:

  1. Navigate to Claude.ai settings

  2. Find the MCP configuration section

  3. Add a new MCP server with:

{
  "name": "directus",
  "command": "npx",
  "args": ["-y", "@staminna/directus-mcp-server"],
  "env": {
    "DIRECTUS_URL": "http://localhost:8065",
    "DIRECTUS_TOKEN": "your-directus-token-here"
  }
}

Note: Claude.ai MCP support may require a Pro subscription and specific browser extensions.


Available Tools

Collection Management

Tool

Description

list_collections

List all collections in Directus

get_collection_schema

Get schema for a specific collection

get_collection_items

Get items from a collection with filtering

create_collection

Create a new collection

delete_collection

Delete a collection (requires confirm)

create_item

Create a new item in a collection

update_item

Update an existing item, optionally into a draft version

delete_items

Delete items by ids or by query (see note below)

bulk_operations

Execute bulk create, update, delete

Schema & Fields

Tool

Description

create_field

Create a new field in a collection

update_field

Update an existing field

delete_field

Delete a field from a collection

create_relationship

Create relationships (O2O, O2M, M2O, M2M, M2A)

analyze_collection_schema

Analyze schema with relationship mapping

validate_collection_schema

Validate schema and relationships

analyze_relationships

Analyze relationships across collections

get_schema_snapshot

Read a full or partial snapshot of the data model

diff_schema

Compare a snapshot against the live schema (merge or mirror). Directus drops request bodies over ~96 KB, so pass a partial snapshot from get_schema_snapshot with include_collections on any sizeable data model — see DIRECTUS_V12_BREAKING_CHANGES.md

apply_schema

Apply a diff (requires confirm)

Flow Management

Tool

Description

get_flows

Get all flows with optional filtering

get_flow

Get a specific flow by ID

create_flow

Create a new automation flow

update_flow

Update an existing flow

delete_flow

Delete a flow

trigger_flow

Manually trigger a flow

get_operations

Get flow operations

User Management

Tool

Description

get_users

Get all users with filtering

get_user

Get a specific user by ID

File Management

Tool

Description

get_files

Get files with filtering and pagination

import_data

Import CSV/JSON into one collection, or several at once

Diagnostics

Tool

Description

diagnose_collection_access

Diagnose collection access issues

refresh_collection_cache

Refresh collection cache

validate_collection_creation

Validate newly created collections

Discovery

Tool

Description

search_tools

Find the tools matching a task description

Tool safety annotations

Every tool carries MCP annotations so a client can tell reads from writes before calling: 17 are readOnlyHint: true, 6 are explicitly destructiveHint: false (additive — creates), and 11 are destructiveHint: true (deletes, overwriting updates, apply_schema, import_data, trigger_flow).

Note that destructiveHint defaults to true in the MCP spec, which is why the additive tools set it to false rather than omitting it.

Deleting items safely

Following Directus 12.3.0, delete_items never falls back to deleting everything:

  • ids: [...] deletes those items.

  • query: {...} deletes everything the query matches.

  • Passing both is rejected.

  • Passing neither deletes nothing and issues no request.

To delete every item in a collection, ask for it explicitly:

{ "collection": "articles", "query": { "limit": -1 }, "confirm": true }

Usage Examples

Once configured, you can interact with Directus through your AI assistant:

"List all collections in my Directus instance"

"Create a new collection called 'blog_posts' with title, content, and published fields"

"Get all items from the 'products' collection where status is 'published'"

"Create a new flow that triggers on item creation in the 'orders' collection"

"Analyze the schema of the 'users' collection including relationships"

Troubleshooting

MCP Server Not Connecting

  1. Verify Directus is running: Ensure your Directus instance is accessible at the configured URL

  2. Check token permissions: The API token needs appropriate permissions for the operations you want to perform

  3. Restart IDE: After changing MCP configuration, fully restart your IDE

  4. Check logs: Look for MCP-related errors in your IDE's developer console

Permission Errors

Ensure your Directus token has the required permissions:

  • Admin token for full access

  • Or configure specific role permissions for collections you need to access

Connection Timeout

If using a remote Directus instance:

  • Verify the URL is correct and accessible

  • Check firewall/network settings

  • Ensure CORS is properly configured on Directus


Development

# Install dependencies
npm install

# Build
npm run build

# Watch mode
npm run dev

# Run server
npm start

# Type check
npm run typecheck

# Lint
npm run lint

Testing

The project ships unit, integration and end-to-end suites (vitest). Coverage thresholds (95% statements/lines/functions/branches) are enforced — the test run fails below them.

# Unit + integration tests
npm test

# With coverage report (coverage/ — text, html, lcov, json-summary)
npm run test:coverage

# End-to-end: builds, then spawns the real server over stdio against a mock Directus
npm run test:e2e

# Everything
npm run test:all

# Refresh the README coverage badges from the last coverage run
npm run badges

Live verification against a real Directus

tests/live/demo.mjs drives all 34 tools against a real instance over stdio. It is deliberately outside npm test — it needs a credential and a reachable server, so it is a manual gate rather than a CI one.

# Read-only + guard phases (touches nothing)
ENV_FILE=.env.mdbaudio npm run test:live

# Also create, mutate and drop a scratch mcp_demo_<stamp> collection
ENV_FILE=.env.mdbaudio npm run test:live -- --write

# Additionally exercise apply_schema, confined to that scratch collection
ENV_FILE=.env.mdbaudio npm run test:live -- --write --apply-schema

Credentials are read from ENV_FILE (default .env.mdbaudio) so they never pass through shell history. Results are reported per tool as pass / refused-by-instance / fail, keeping "this server is broken" separate from "this instance declined". --apply-schema diffs in merge mode, which yields a strictly additive diff, so it can only re-create the scratch collection — it cannot drop anything that already existed. Cleanup runs even when an earlier phase fails.

The e2e suite uses the official MCP SDK client (StdioClientTransport) to spawn dist/index.js as a subprocess, talking to an in-process mock Directus on an ephemeral port — no real Directus instance or network access needed.


Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

  1. Fork the repository

  2. Create your feature branch (git checkout -b feature/amazing-feature)

  3. Commit your changes (git commit -m 'Add some amazing feature')

  4. Push to the branch (git push origin feature/amazing-feature)

  5. Open a Pull Request


License

MIT © Jorge Domingues Nunes


Related MCP Servers

  • F
    license
    Not graded
    quality
    Not graded
    maintenance
    Enables comprehensive management of Storyblok CMS through natural language interactions. Supports story creation and publishing, asset management, component schema updates, release workflows, and content discovery across all major Storyblok APIs.
    10
  • A
    license
    A
    quality
    C
    maintenance
    Enables comprehensive management of Directus instances through tools for schema manipulation, content CRUD operations, and dashboard management. It allows AI assistants to programmatically interact with collections, fields, relations, and workflow automation using the official Directus SDK.
    20
    40
    MIT

View all related MCP servers

Related MCP Connectors

  • Manage Appwrite projects, databases, auth, storage, functions, and messaging; search Appwrite docs

  • AI-powered design and management for Webflow Sites

  • Manage SRG+ hubs, channels, content, assets, users, and workspaces from any MCP-aware AI agent.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/staminna/mcp-server-claude'

If you have feedback or need assistance with the MCP directory API, please join our Discord server