Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Python SDK for building AI applications on Azure AI Foundry. Manages agents, threads, evaluations, connections, deployments, datasets, and indexes through a single project client. Supports both native Foundry operations and an OpenAI-compatible client interface.
Provides unified access to agents, evaluations, connections, and deployments through a single Foundry-aware client rather than managing multiple Azure SDK clients separately.
When to use it
- Building versioned customer support agents for production deployments
- Running fluency and task-adherence evaluations on agent outputs
- Searching Azure AI indexes from within an agent workflow
- Listing and selecting model deployments across a Foundry project
- Creating code-executing agents that run Python on user requests
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: 'Set up an agent that can execute code and answer questions'
- 1Authenticate with DefaultAzureCredential and initialize AIProjectClient
- 2Create an agent with CodeInterpreterTool and custom instructions
- 3Open a thread and post the user message
- 4Call create_and_process to run the agent on the thread
- 5Read assistant messages from the completed run
Agent response with executed code results returned from the thread messages
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.