How to Use the Directus MCP in Pydantic AI

Q: How do I connect Directus to Pydantic AI?

Install the package with the MCP extra, then initialize an MCPToolset with your HTTP endpoint. Pass this toolset into the Agent constructor using the toolsets parameter to expose all ten tools.

Q: Can I run the Directus MCP Server locally with Pydantic AI?

Yes. You can run the server locally and connect using the MCPToolset HTTP client, or let Vinkius host the server and connect securely using your hosted endpoint token in your Python code.

Q: How do I handle database deletions safely?

You should wrap the wipecmsrecord tool in a custom validation step or restrict its access entirely. This ensures that your agent cannot delete rows without meeting your strict Pydantic model constraints.

Run type-safe Pydantic AI agents that validate Directus database schemas and records at runtime using this MCP Server.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Directus MCP to Pydantic AI

Create your Vinkius account to connect Directus to Pydantic AI and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Directus with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Enforce schema validation with this MCP Server

Stop letting agents guess your database structure and cause runtime crashes. Your agent uses `list_schema_collections` and `list_collection_fields` to inspect your tables and columns, returning strict schemas that map directly to Pydantic models. This integration ensures that any data returned by `list_collection_items` is validated against your Python type definitions before your application processes it. If the database returns unexpected nulls or modified types, the framework fails loudly.

Validated record mutations

Write to your database with complete confidence. The agent uses `create_cms_record` and `patch_cms_record` to insert or update rows, validating the JSON payloads against your Pydantic schemas before executing the write. This prevents corrupt data from entering your tables. If an agent attempts an invalid update or tries to call `wipe_cms_record` with the wrong ID, the validation layer catches the error before any database damage occurs.

Administrative and file audits

Manage your media assets and user tables using type-safe queries. The agent calls `list_directus_files` to audit media storage paths and `list_directus_users` to verify active admin accounts against strict Python models. For granular checks, `get_single_item` retrieves specific database rows by UUID. This gives your agent a reliable, validated view of your entire CMS state without silent errors or hallucinated fields.

Setup guide

Set up Directus MCP in Pydantic AI

Prerequisites

Python 3.10+ installed
pydantic-ai-slim[fastmcp] package
Active Vinkius subscription with a valid endpoint token

1

Install Pydantic AI with FastMCP
Run pip install "pydantic-ai-slim[fastmcp]". The FastMCP toolset replaces the deprecated MCPServerHTTP class with full protocol support.
2

Configure the FastMCPToolset
Pass a JSON-style config dict to FastMCPToolset with your Vinkius URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. Supports Streamable HTTP, SSE, and Stdio transports.
3

Create and run your agent
Pass the toolset to Agent(toolsets=[toolset]) and call agent.run(). Swap openai:gpt-4o for any supported model — Anthropic, Google, Mistral, or Groq.

agent.py

from pydantic_ai import Agent
from pydantic_ai.toolsets.fastmcp import FastMCPToolset

toolset = FastMCPToolset({
    "mcpServers": {
        "directus-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

agent = Agent(
    "openai:gpt-4o",
    toolsets=[toolset],
    system_prompt="You have access to Directus tools.",
)

result = await agent.run("List recent Directus transactions")
print(result.output)

Get your connection token →

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.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Directus now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Directus MCP in Pydantic AI

Install the package with the MCP extra, then initialize an `MCPToolset` with your HTTP endpoint. Pass this toolset into the `Agent` constructor using the `toolsets` parameter to expose all ten tools.

The framework will immediately raise a validation error at runtime. This prevents your agent from working with corrupted data or hallucinating fields that do not exist in your SQL tables.

Yes. You can run the server locally and connect using the `MCPToolset` HTTP client, or let Vinkius host the server and connect securely using your hosted endpoint token in your Python code.

You should wrap the `wipe_cms_record` tool in a custom validation step or restrict its access entirely. This ensures that your agent cannot delete rows without meeting your strict Pydantic model constraints.

Yes, all data traffic passes through secure, isolated V8 sandboxes. Your user records, file metadata, and SQL database rows are processed in-memory and are never stored or logged by Vinkius.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript