AgentSkillsCN

azure-functions

帮助使用 Python v2 编程模型创建、调试并部署 Azure Functions。适用于处理 HTTP 触发器、MCP 触发器、定时器触发器,或任何 Azure Functions 开发工作,包括使用 func start 和 Azurite 进行本地测试。

SKILL.md
--- frontmatter
name: azure-functions
description: Helps create, debug, and deploy Azure Functions using the Python v2 programming model. Use this skill when working with HTTP triggers, MCP triggers, timer triggers, or any Azure Functions development including local testing with func start and Azurite.

Azure Functions Development Skill

Overview

This skill helps you build Azure Functions using the Python v2 programming model with decorators.

Key Patterns

HTTP Trigger Function

python
import azure.functions as func
import json

app = func.FunctionApp(http_auth_level=func.AuthLevel.FUNCTION)

@app.function_name(name="my_function")
@app.route(route="api/endpoint", methods=["GET", "POST"])
def my_function(req: func.HttpRequest) -> func.HttpResponse:
    """HTTP trigger function with proper error handling."""
    try:
        # Parse request
        data = req.get_json() if req.method == "POST" else dict(req.params)
        
        # Your logic here
        result = {"status": "success", "data": data}
        
        return func.HttpResponse(
            body=json.dumps(result),
            status_code=200,
            mimetype="application/json"
        )
    except Exception as e:
        return func.HttpResponse(
            body=json.dumps({"error": str(e)}),
            status_code=500,
            mimetype="application/json"
        )

MCP Tool Trigger

python
@app.generic_trigger(
    arg_name="context",
    type="mcpToolTrigger",
    toolName="my_tool",
    description="What this tool does and when to use it",
    toolProperties='[{"propertyName": "input", "propertyType": "string", "description": "The input parameter"}]'
)
def my_tool(context) -> str:
    """MCP tool function."""
    content = json.loads(context)
    input_value = content["arguments"].get("input", "")
    return json.dumps({"result": input_value})

Configuration Files

host.json (Runtime Config)

json
{
  "version": "2.0",
  "extensionBundle": {
    "id": "Microsoft.Azure.Functions.ExtensionBundle",
    "version": "[4.*, 5.0.0)"
  },
  "extensions": {
    "mcp": {
      "enabled": true
    }
  }
}

local.settings.json (Local Development)

json
{
  "IsEncrypted": false,
  "Values": {
    "FUNCTIONS_WORKER_RUNTIME": "python",
    "AzureWebJobsStorage": "UseDevelopmentStorage=true",
    "AzureWebJobsFeatureFlags": "EnableWorkerIndexing"
  }
}

Local Development Commands

bash
# Start Azurite storage emulator (required)
azurite --silent --location /tmp/azurite

# Start the function app locally
cd src/function_app_directory
func start

# Install Azure Functions Core Tools (macOS)
brew tap azure/functions
brew install azure-functions-core-tools@4

Deployment with Azure Developer CLI

bash
# Initialize project from template
azd init --template <template-name> -e <environment-name>

# Deploy to Azure
azd up

# Get function keys
az functionapp keys list --resource-group <rg> --name <app-name>

Common Issues and Solutions

IssueSolution
AzureWebJobsStorage errorStart Azurite: azurite --silent
Functions not discoveredEnsure host.json has correct extension bundle
MCP tools not showingCheck mcp.enabled: true in host.json extensions
401 UnauthorizedInclude x-functions-key header for FUNCTION auth level

Best Practices

  1. Always use type hints for better IDE support
  2. Log important operations with logging.info() and logging.error()
  3. Return JSON responses with appropriate status codes
  4. Use FUNCTION auth level for production (not ANONYMOUS)
  5. Test locally with func start before deploying