MCP For Claude Marketplace
Azure Cosmos DB Container

Azure Cosmos DB Container MCP Connector for Claude

F

This MCP does exactly one thing: it manages documents in a single Azure Cosmos DB Container. That's its only function, and nothing else. Incredible for giving your AI a secure NoSQL database.

4 tools Official Updated Jun 28, 2026 Official Vinkius Partner
More Details Connect to Claude

This server strips away dangerous global Azure permissions. It gives your AI agent one surgical superpower: the ability to query, insert, and update documents inside one specific Cosmos DB Container.

By strictly scoping access, your AI can safely manage structured data, store chat histories, and process complex NoSQL queries without ever touching your critical cloud databases.

The Superpowers

  • Absolute Containment: The agent is locked to a single container. It cannot list other databases or drop your production data.
  • Native Cosmos DB Integration: Direct interactions with Cosmos DB, supporting rich SQL queries and partition management.
  • Plug & Play Database: Instantly gives your agent a scalable NoSQL database to store structured memories and application state.
nosqldocument-databasedata-storagescoped-accesscloud-databasejson-storage

4 tools expose this connector's capabilities to your AI agent.

create_document

Make sure to provide the ID and Partition Key properties inside the documentJson if required. Create a new document in the Cosmos DB Container

delete_document

Provide partitionKey if your container requires it. Delete a document from the Cosmos DB Container

get_document

Provide partitionKey if your container requires it. Retrieve a specific document by its ID

query_documents

You can optionally provide parameters in JSON format. Do not include the DB or Container name in the query, Cosmos expects queries like "SELECT * FROM c WHERE c.status = @status". Execute a SQL query against the configured Cosmos DB Container

See how to talk to your AI agent using Azure Cosmos DB Container.

Query the database for all documents where 'status' is 'pending'.

I executed `SELECT * FROM c WHERE c.status = 'pending'`. I found 12 pending documents.

Create a new document with id 'user_123' and name 'John Doe'.

The document was successfully created in the Cosmos DB container.

Delete the document with id 'old_report_456'.

I have permanently deleted the document 'old_report_456' from the database.

To enforce zero-trust security. An autonomous AI agent should not have the ability to read or delete data across your entire Cosmos DB Account (which might include user passwords, financial records, etc.). By scoping it to a single container, you eliminate the risk of catastrophic data loss.

Related Connectors

RI

Reddit Intelligence MCP

9 tools Official

Tap into Reddit's public communities for real-time marketing intelligence — discover trending topics, monitor niches, and analyze sentiment with a free Reddit app.

A+ View details →
AS

Asaas MCP

8 tools Official

Manage payments, customers, and subscriptions with Asaas — the complete digital account for Brazilian businesses via AI.

D View details →
DA

Data.gov MCP

13 tools Official

Search 300,000+ US government open datasets — agriculture, climate, education, health, finance and more.

A+ View details →
MA

Marchex MCP

10 tools Official

Marchex Conversation Intelligence — track calls, manage campaigns, and analyze performance.

A+ View details →