DirectusSAFE
MCP server for Directus API integration
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://smithery.ai/server/@pixelsock/directus-mcp)
A Node.js server implementing Model Context Protocol (MCP) for Directus CMS. Enable AI Clients to interact with the Directus API through the Model Context Protocol (MCP).
i Prerequisites
▶️ Quick start
- Get your Directus API credentials
- Go to your Directus instance
- Create a static access token or get your email and password
- Keep these credentials secure
- Add to your AI editor
{
"mcpServers": {
"directus": {
"command": "npx",
"args": ["-y", "@pixelsock/directus-mcp@latest"],
"env": {
"DIRECTUS_URL": "https://your-directus-instance.com",
"DIRECTUS_ACCESS_TOKEN": "YOUR_ACCESS_TOKEN"
}
}
}
}Alternatively, you can use email/password authentication:
{
"mcpServers": {
"directus": {
"command": "npx",
"args": ["-y", "@pixelsock/directus-mcp@latest"],
"env": {
"DIRECTUS_URL": "https://your-directus-instance.com",
"DIRECTUS_EMAIL": "[email protected]",
"DIRECTUS_PASSWORD": "your-password"
}
}
}
}For Cursor:
- Go to Settings → Cursor Settings → MCP
- Click
+ Add New Global MCP Server - Paste configuration
- Replace placeholder values with your Direct
4b08a1c6ffedOBSERVED · 2026-10-09Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add directus-api-extended --env DIRECTUS_ACCESS_TOKEN=${DIRECTUS_ACCESS_TOKEN} --env DIRECTUS_PASSWORD=${DIRECTUS_PASSWORD} -- npx -y [email protected]{
"mcpServers": {
"directus-api-extended": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"DIRECTUS_ACCESS_TOKEN": "${DIRECTUS_ACCESS_TOKEN}",
"DIRECTUS_PASSWORD": "${DIRECTUS_PASSWORD}"
}
}
}
}Exposed tools (18)
14 read · 3 write · 1 destructive. Blast radius: 1 tool can delete or overwrite — an agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
createItem | write | Create a new item in a collection |
deleteItem | destructive | Delete an item from a collection |
getActivity | read | Get activity logs from Directus |
getCollections | read | Get all collection schemas from Directus |
getConfig | read | Get current configuration information (without secrets) |
getCurrentUser | read | Get the current user info |
getFields | read | Get fields for a collection |
getFiles | read | Get files from Directus |
getItem | read | Get a single item from a collection by ID |
getItems | read | Get items from a collection in Directus |
getPermissions | read | Get permissions from Directus |
getRelations | read | Get relations for a collection |
getRoles | read | Get roles from Directus |
getSystemInfo | read | Get system information from Directus |
getUsers | read | Get users from Directus |
login | read | Login to Directus and get an access token |
updateItem | write | Update an existing item in a collection |
uploadFile | write | Upload a file to Directus |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (5 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (5)
deleteItem
@modelcontextprotocol/sdk, axios, form-data, zod, @modelcontextprotocol/inspector, @types/node, concurrently, nodemon
- Create a static access token or get your email and password
2. Check that your access token or email/password credentials are valid
Gates applied: no_behavioural_pass, no_license.
4b08a1c6ffedfull audit observations/trust-audit/mcp-server/pixelsock__directus-1.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-09 | 4b08a1c6ffed | SAFE | B | 89 | first audit |
Questions
What is the Directus MCP server?
MCP server for Directus API integration
What tools does Directus expose?
18 in total: 14 read-only, 3 that write, and 1 that can delete or overwrite (deleteItem). Every one is listed on this page with its risk.
Is Directus safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B. Separately from the audit: 1 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does Directus need?
It reads DIRECTUS_ACCESS_TOKEN and DIRECTUS_PASSWORD from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does Directus run?
It speaks stdio and streamable-http, so it runs as a local process your client starts. It is published on npm as directus-api-extended at 1.0.0.
How current is this page?
The grade is for one exact copy of the source (4b08a1c6ffed), read on 2026-10-09. The repository is watched and re-audited when it changes.