azure-cosmos-py

What This Skill Does

The azure-cosmos-py skill provides an interface to the official Azure Cosmos DB Python SDK, enabling AI agents to manage NoSQL database resources programmatically. This skill allows for seamless CRUD (Create, Read, Update, Delete) operations on documents, management of databases and containers, and the execution of SQL-like queries across globally distributed data. It abstracts the complexity of authentication via Azure Identity, allowing users to perform cloud-native data operations directly through their AI agent.

Installation

To integrate this skill, use the clawhub installation command:

clawhub install openclaw/skills/skills/thegovind/azure-cosmos-py

Ensure your local environment or AI runtime has access to Azure credentials, typically via environment variables (COSMOS_ENDPOINT) and the azure-identity package.

Use Cases

This skill is ideal for:

• Data Engineering: Automating the ingestion or migration of records into Azure Cosmos DB containers.
• Application Backend Orchestration: Building dynamic AI agents that can read, update, and manage state in a NoSQL database in real-time.
• Analytics Assistance: Fetching and filtering large datasets using complex SQL queries for downstream processing or summarizing.
• Cloud Operations: Provisioning and teardown of development databases and containers during CI/CD pipelines.

Example Prompts

1. "Look up the document with ID 'user-882' in the 'profiles' container where the partition key is 'us-east' and update the 'last_login' field to today's date."
2. "Query the 'orders' container to return all items where the price is greater than 1000 and the category is 'electronics'."
3. "Create a new container named 'sessions' with a partition key of '/session_id' inside my existing database."

Tips & Limitations

• Partition Key Awareness: Always ensure you include the partition key in your queries and CRUD operations. Operations without a partition key may perform poorly or require a cross-partition query, which is computationally expensive.
• Security: Never hardcode your credentials. Use managed identities or Environment Variables like COSMOS_ENDPOINT to maintain a secure posture.
• Rate Limiting: Be mindful of Request Units (RU) assigned to your Cosmos DB instance. High-frequency operations via this skill may consume your throughput quickly, potentially leading to 429 'Too Many Requests' errors.