mcpbeat Sign in

Azure Patterns Agent Skill

Azure Functions, Cosmos DB modeling, Service Bus patterns, Bicep templates

2k tokens
context cost
the whole folder, loaded on every use
1
files
instructions only
0
copies elsewhere
how many repositories repackaged it
521
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/vibeeval/vibecosystem --skill azure-patterns

The instruction itself

9 sections, as written by the author

Azure Patterns

Azure Functions

HTTP Trigger (Node.js v4)

import { app, HttpRequest, HttpResponseInit, InvocationContext } from '@azure/functions';

app.http('getOrder', {
  methods: ['GET'],
  authLevel: 'function',
  route: 'orders/{orderId}',
  handler: async (request: HttpRequest, context: InvocationContext): Promise<HttpResponseInit> => {
    const orderId = request.params.orderId;
    context.log(`Processing order: ${orderId}`);

    try {
      const order = await orderService.getById(orderId);
      if (!order) {
        return { status: 404, jsonBody: { error: 'Order not found' } };
      }
      return { status: 200, jsonBody: order };
    } catch (error) {
      context.error('Failed to get order', error);
      return { status: 500, jsonBody: { error: 'Internal server error' } };
    }
  },
});

// Service Bus trigger with retry
app.serviceBusTopic('processOrderEvent', {
  topicName: 'order-events',
  subscriptionName: 'order-processor',
  connection: 'ServiceBusConnection',
  handler: async (message: unknown, context: InvocationContext) => {
    const event = message as OrderEvent;
    context.log(`Processing event: ${event.type} for order ${event.orderId}`);

    await processEvent(event);
  },
});

Durable Functions (Orchestrator)

import * as df from 'durable-functions';

df.app.orchestration('orderWorkflow', function* (context) {
  const orderId = context.df.getInput() as string;

  // Step 1: Validate order
  const order = yield context.df.callActivity('validateOrder', orderId);

  // Step 2: Reserve inventory (with retry)
  const retryOptions = new df.RetryOptions(5000, 3); // 5s interval, 3 attempts
  yield context.df.callActivityWithRetry('reserveInventory', retryOptions, order);

  // Step 3: Charge payment
  yield context.df.callActivity('chargePayment', order);

  // Step 4: Wait for shipping confirmation (with timeout)
  const deadline = new Date(context.df.currentUtcDateTime.getTime() + 24 * 60 * 60 * 1000);
  const shippingEvent = context.df.waitForExternalEvent('shippingConfirmed');
  const timeout = context.df.createTimer(deadline);

  const winner = yield context.df.Task.any([shippingEvent, timeout]);
  if (winner === timeout) {
    yield context.df.callActivity('escalateShipping', orderId);
  }

  return { orderId, status: 'completed' };
});

Cosmos DB Modeling

// Partition key design: use tenant/user ID for multi-tenant
interface OrderDocument {
  id: string;                    // Unique document ID
  partitionKey: string;          // = tenantId (good distribution)
  type: 'order';                 // Discriminator for polymorphic container
  customerId: string;
  items: OrderItem[];            // Embed frequently accessed together
  total: number;
  status: string;
  createdAt: string;
  _etag?: string;                // Optimistic concurrency
}

// Optimistic concurrency with ETags
async function updateOrder(order: OrderDocument): Promise<void> {
  const container = cosmosClient.database('shop').container('orders');
  try {
    await container.item(order.id, order.partitionKey).replace(order, {
      accessCondition: { type: 'IfMatch', condition: order._etag },
    });
  } catch (error) {
    if (error.code === 412) {
      throw new ConflictError('Order was modified by another process');
    }
    throw error;
  }
}

// Change feed processor for event-driven updates
const changeFeedProcessor = cosmosClient
  .database('shop')
  .container('orders')
  .getChangeFeedProcessorBuilder('orderProcessor')
  .setFeedProcessorOptions({ startFromBeginning: false, maxItemCount: 25 })
  .setLeaseContainer(leaseContainer)
  .setProcessChanges(async (changes, context) => {
    for (const doc of changes) {
      await publishEvent({ type: 'order.updated', data: doc });
    }
  })
  .build();

Service Bus Patterns

import { ServiceBusClient, ServiceBusMessage } from '@azure/service-bus';

const client = new ServiceBusClient(process.env.SERVICEBUS_CONNECTION!);

// Send with deduplication and scheduling
async function sendOrderEvent(event: OrderEvent): Promise<void> {
  const sender = client.createSender('order-events');
  const message: ServiceBusMessage = {
    body: event,
    messageId: `${event.orderId}-${event.type}-${Date.now()}`,  // Dedup key
    subject: event.type,
    applicationProperties: { priority: event.priority },
    timeToLive: 24 * 60 * 60 * 1000,  // 24h TTL
  };
  await sender.sendMessages(message);
  await sender.close();
}

// Receive with sessions (ordered processing per entity)
async function processWithSessions(): Promise<void> {
  const receiver = client.acceptSession('order-events', 'order-processor', 'session-1');
  const messages = await receiver.receiveMessages(10, { maxWaitTimeInMs: 5000 });

  for (const msg of messages) {
    try {
      await handleMessage(msg.body);
      await receiver.completeMessage(msg);
    } catch {
      await receiver.abandonMessage(msg);  // Retry via Service Bus
    }
  }
  await receiver.close();
}

Bicep Templates

@description('Azure Function App with Service Bus and Cosmos DB')
param location string = resourceGroup().location
param appName string

resource cosmosAccount 'Microsoft.DocumentDB/databaseAccounts@2024-05-15' = {
  name: '${appName}-cosmos'
  location: location
  kind: 'GlobalDocumentDB'
  properties: {
    databaseAccountOfferType: 'Standard'
    consistencyPolicy: { defaultConsistencyLevel: 'Session' }
    locations: [{ locationName: location, failoverPriority: 0 }]
  }
}

resource serviceBus 'Microsoft.ServiceBus/namespaces@2022-10-01-preview' = {
  name: '${appName}-bus'
  location: location
  sku: { name: 'Standard', tier: 'Standard' }
}

resource functionApp 'Microsoft.Web/sites@2023-12-01' = {
  name: '${appName}-func'
  location: location
  kind: 'functionapp,linux'
  identity: { type: 'SystemAssigned' }
  properties: {
    siteConfig: {
      appSettings: [
        { name: 'CosmosDBConnection', value: cosmosAccount.listConnectionStrings().connectionStrings[0].connectionString }
        { name: 'ServiceBusConnection', value: listKeys('${serviceBus.id}/AuthorizationRules/RootManageSharedAccessKey', serviceBus.apiVersion).primaryConnectionString }
      ]
    }
  }
}

Checklist

  • [ ] Functions use managed identity, not connection strings where possible
  • [ ] Cosmos DB partition key chosen for even distribution and query patterns
  • [ ] Service Bus dead-letter queue monitored and processed
  • [ ] Durable Functions used for multi-step workflows
  • [ ] Bicep templates parameterized for multi-environment deployment
  • [ ] Cosmos DB RU autoscale configured to handle traffic spikes
  • [ ] Application Insights connected for distributed tracing
  • [ ] Function timeout matches expected workload duration

Anti-Patterns

  • Cross-partition queries in Cosmos DB (expensive, slow)
  • Storing large blobs in Cosmos DB (use Blob Storage + reference)
  • Not using sessions for ordered message processing
  • Shared Service Bus connection strings across services
  • Function apps without Application Insights (blind to failures)
  • Cosmos DB with fixed RU provisioning for variable workloads
  • Not configuring dead-letter queues on Service Bus subscriptions
  • Using host keys for Function auth (use managed identity + RBAC)

Other skills for the same job

different authors, same section of the catalogue
Azure Kubernetes Automatic Readiness
by microsoft
vendor ×3

Assess Kubernetes workloads and cluster configuration for AKS Automatic compatibility. Identifies incompatibilities, generates fixes, and guides migration from AKS Standard to AKS Automatic. WHEN: migrate to AKS Automatic, check AKS Automatic readiness, validate manifests for Automatic, assess cluster for Automatic compatibility, fix deployment for Automatic compatibility, identify AKS Automatic migration blockers, is my cluster ready for AKS Automatic.

13k tokens
Capacity
by microsoft
vendor ×3

Discovers available Azure OpenAI model capacity across regions and projects. Analyzes quota limits, compares availability, and recommends optimal deployment locations based on capacity requirements. USE FOR: find capacity, check quota, where can I deploy, capacity discovery, best region for capacity, multi-project capacity search, quota analysis, model availability, region comparison, check TPM availability. DO NOT USE FOR: actual deployment (hand off to preset or customize after discovery), quota increase requests (direct user to Azure Portal), listing existing deployments.

6k tokens scripts
Customize
by microsoft
vendor ×3

Interactive guided deployment flow for Azure OpenAI models with full customization control. Step-by-step selection of model version, SKU (GlobalStandard/Standard/ProvisionedManaged), capacity, RAI policy (content filter), and advanced options (dynamic quota, priority processing, spillover). USE FOR: custom deployment, customize model deployment, choose version, select SKU, set capacity, configure content filter, RAI policy, deployment options, detailed deployment, advanced deployment, PTU deployment, provisioned throughput. DO NOT USE FOR: quick deployment to optimal region (use preset).

8k tokens
Deploy Model
by microsoft
vendor ×3

Unified Azure OpenAI model deployment skill with intelligent intent-based routing. Handles quick preset deployments, fully customized deployments (version/SKU/capacity/RAI policy), and capacity discovery across regions and projects. USE FOR: deploy model, deploy gpt, create deployment, model deployment, deploy openai model, set up model, provision model, find capacity, check model availability, where can I deploy, best region for model, capacity analysis. DO NOT USE FOR: listing existing deployments (use foundry_models_deployments_list MCP tool), deleting deployments, agent creation (use agent/create), project creation (use project/create).

26k tokens scripts
Preset
by microsoft
vendor ×3

Intelligently deploys Azure OpenAI models to optimal regions by analyzing capacity across all available regions. Automatically checks current region first and shows alternatives if needed. USE FOR: quick deployment, optimal region, best region, automatic region selection, fast setup, multi-region capacity check, high availability deployment, deploy to best location. DO NOT USE FOR: custom SKU selection (use customize), specific version selection (use customize), custom capacity configuration (use customize), PTU deployments (use customize).

9k tokens
Lamindb
by christophacham
×3

This skill should be used when working with LaminDB, an open-source data framework for biology that makes data queryable, traceable, reproducible, and FAIR. Use when managing biological datasets (scRNA-seq, spatial, flow cytometry, etc.), tracking computational workflows, curating and validating data with biological ontologies, building data lakehouses, or ensuring data lineage and reproducibility in biological research. Covers data management, annotation, ontologies (genes, cell types, diseases, tissues), schema validation, integrations with workflow managers (Nextflow, Snakemake) and MLOps platforms (W&B, MLflow), and deployment strategies.

22k tokens
Latchbio Integration
by christophacham
×3

Latch platform for bioinformatics workflows. Build pipelines with Latch SDK, @workflow/@task decorators, deploy serverless workflows, LatchFile/LatchDir, Nextflow/Snakemake integration.

12k tokens
Modal
by christophacham
×3

Run Python code in the cloud with serverless containers, GPUs, and autoscaling. Use when deploying ML models, running batch processing jobs, scheduling compute-intensive tasks, or serving APIs that require GPU acceleration or dynamic scaling.

17k tokens

How to use it

Copy the folder

Take vibeeval/azure-patterns from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.