agent-framework-azure-ai-py
Build Azure AI Foundry Agents
Create persistent AI agents on Azure AI Foundry using the Microsoft Agent Framework Python SDK. This skill provides code examples for building agents with hosted tools, function tools, and conversation threads.
スキルZIPをダウンロード
Claudeでアップロード
設定 → 機能 → スキル → スキルをアップロードへ移動
オンにして利用開始
テストする
「agent-framework-azure-ai-py」を使用しています。 Create a basic Azure AI agent
期待される結果:
The agent will be created with AzureAIAgentsProvider. It can then process user queries and return responses using the configured instructions and tools.
「agent-framework-azure-ai-py」を使用しています。 What's the weather in Seattle?
期待される結果:
Weather in Seattle: 72°F, sunny
「agent-framework-azure-ai-py」を使用しています。 Calculate factorial of 20
期待される結果:
The agent uses HostedCodeInterpreterTool to execute Python code and return the result: 2432902008176640000
セキュリティ監査
安全All 57 static findings are false positives. The skill contains only documentation and code examples for building Azure AI Foundry agents using the Microsoft Agent Framework Python SDK. The detected patterns (import statements, async context managers, markdown backticks, placeholder URLs) are legitimate documentation elements, not security risks.
リスク要因
⚡ スクリプトを含む (9)
⚙️ 外部コマンド (40)
🌐 ネットワークアクセス (2)
品質スコア
作れるもの
Build Customer Service Agents
Create persistent AI agents that maintain conversation context across multiple interactions
Research Assistant with Tools
Build agents that can search the web, execute code, and analyze results
Enterprise Workflow Automation
Create agents that integrate with enterprise systems via MCP tools
これらのプロンプトを試す
Create an Azure AI Foundry agent named 'MyAgent' with the instruction 'You are a helpful assistant' using the agent-framework-azure-ai-py skill.
Create an agent that can get weather information. Define a get_weather function that takes a location parameter and returns weather data.
Build an agent using HostedCodeInterpreterTool and HostedWebSearchTool that can execute Python code and search the web.
Create an agent that maintains conversation context across multiple turns using thread management.
ベストプラクティス
- Always use async context managers (async with) for proper resource cleanup
- Pass functions directly to the tools parameter - they are auto-converted to AIFunction
- Use Annotated[type, Field(description=...)] for function parameters to provide clear descriptions
- Use get_new_thread() for multi-turn conversations to maintain context
回避
- Do not use synchronous operations inside async functions - this will cause deadlocks
- Avoid hardcoding credentials - use AzureCliCredential or DefaultAzureCredential
- Do not skip the async context manager - always use 'async with' for providers
- Avoid passing non-async functions directly without proper wrapping
よくある質問
What is Azure AI Foundry?
Do I need an Azure subscription?
What authentication methods are supported?
Can I use Claude Code with Azure AI agents?
What are hosted tools?
How do I manage conversation state?
開発者の詳細
作成者
sickn33ライセンス
MIT
リポジトリ
https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/agent-framework-azure-ai-py参照
main
ファイル構成
📄 SKILL.md