# SkillPatch skill: azure-servicebus-ts

This skill provides structured guidance for building enterprise messaging applications using the Azure Service Bus SDK for TypeScript (@azure/service-bus). It covers authentication, sending and receiving messages, topics/subscriptions, message sessions, and dead-letter handling. Agents can follow the included workflows and TypeScript code examples to implement queues, event-driven messaging, and session-based patterns.

You (the agent) were given this URL and asked to use this skill. This is a **public** skill — no SkillPatch account, API key, or prior setup is required. Two ways to use it:

**1. Use it right now** — the skill's complete file tree (SKILL.md + all reference files) is inlined below; read `SKILL.md` first, then follow it, consulting the other files as it directs.

**2. Install the exact package onto disk** (recommended if you can run a shell — this reproduces the skill byte-for-byte, including any binary assets that can't be inlined):

```bash
mkdir -p .claude/skills/azure-servicebus-ts
curl -sSL https://skillpatch.dev/install_skill/azure-servicebus-ts | tar -xz -C .claude/skills/
```

(`.claude/skills/` is Claude Code's convention; use whatever directory your agent loads skills from.)


---

## Skill files (3)

- `SKILL.md`
- `references/error-handling.md`
- `references/queues-topics.md`


### `SKILL.md`

````markdown
---
name: azure-servicebus-ts
description: Build messaging applications using Azure Service Bus SDK for JavaScript (@azure/service-bus). Use when implementing queues, topics/subscriptions, message sessions, dead-letter handling, or enterprise messaging patterns.
license: MIT
metadata:
  author: Microsoft
  version: "1.0.0"
  package: '@azure/service-bus'
---

# Azure Service Bus SDK for TypeScript

Enterprise messaging with queues, topics, and subscriptions.

## Installation

```bash
npm install @azure/service-bus @azure/identity
```

## Environment Variables

```bash
SERVICEBUS_NAMESPACE=<namespace>.servicebus.windows.net
SERVICEBUS_QUEUE_NAME=my-queue
SERVICEBUS_TOPIC_NAME=my-topic
SERVICEBUS_SUBSCRIPTION_NAME=my-subscription
AZURE_TOKEN_CREDENTIALS=prod # Required only if DefaultAzureCredential is used in production
```

## Authentication

```typescript
import { ServiceBusClient } from "@azure/service-bus";
import { DefaultAzureCredential, ManagedIdentityCredential } from "@azure/identity";

// Local dev: DefaultAzureCredential. Production: set AZURE_TOKEN_CREDENTIALS=prod or AZURE_TOKEN_CREDENTIALS=<specific_credential>
const credential = new DefaultAzureCredential({requiredEnvVars: ["AZURE_TOKEN_CREDENTIALS"]});
// Or use a specific credential directly in production:
// See https://learn.microsoft.com/javascript/api/overview/azure/identity-readme?view=azure-node-latest#credential-classes
// const credential = new ManagedIdentityCredential();

const fullyQualifiedNamespace = process.env.SERVICEBUS_NAMESPACE!;
const client = new ServiceBusClient(fullyQualifiedNamespace, credential);
```

## Core Workflow

### Send Messages to Queue

```typescript
const sender = client.createSender("my-queue");

// Single message
await sender.sendMessages({
  body: { orderId: "12345", amount: 99.99 },
  contentType: "application/json",
});

// Batch messages
const batch = await sender.createMessageBatch();
batch.tryAddMessage({ body: "Message 1" });
batch.tryAddMessage({ body: "Message 2" });
await sender.sendMessages(batch);

await sender.close();
```

### Receive Messages from Queue

```typescript
const receiver = client.createReceiver("my-queue");

// Receive batch
const messages = await receiver.receiveMessages(10, { maxWaitTimeInMs: 5000 });
for (const message of messages) {
  console.log(`Received: ${message.body}`);
  await receiver.completeMessage(message);
}

await receiver.close();
```

### Subscribe to Messages (Event-Driven)

```typescript
const receiver = client.createReceiver("my-queue");

const subscription = receiver.subscribe({
  processMessage: async (message) => {
    console.log(`Processing: ${message.body}`);
    // Message auto-completed on success
  },
  processError: async (args) => {
    console.error(`Error: ${args.error}`);
  },
});

// Stop after some time
setTimeout(async () => {
  await subscription.close();
  await receiver.close();
}, 60000);
```

### Topics and Subscriptions

```typescript
// Send to topic
const topicSender = client.createSender("my-topic");
await topicSender.sendMessages({
  body: { event: "order.created", data: { orderId: "123" } },
  applicationProperties: { eventType: "order.created" },
});

// Receive from subscription
const subscriptionReceiver = client.createReceiver("my-topic", "my-subscription");
const messages = await subscriptionReceiver.receiveMessages(10);
```

## Message Sessions

```typescript
// Send session message
const sender = client.createSender("session-queue");
await sender.sendMessages({
  body: { step: 1, data: "First step" },
  sessionId: "workflow-123",
});

// Receive session messages
const sessionReceiver = await client.acceptSession("session-queue", "workflow-123");
const messages = await sessionReceiver.receiveMessages(10);

// Get/set session state
const state = await sessionReceiver.getSessionState();
await sessionReceiver.setSessionState(Buffer.from(JSON.stringify({ progress: 50 })));

await sessionReceiver.close();
```

## Dead-Letter Handling

```typescript
// Move to dead-letter
await receiver.deadLetterMessage(message, {
  deadLetterReason: "Validation failed",
  deadLetterErrorDescription: "Missing required field: orderId",
});

// Process dead-letter queue
const dlqReceiver = client.createReceiver("my-queue", { subQueueType: "deadLetter" });
const dlqMessages = await dlqReceiver.receiveMessages(10);
for (const msg of dlqMessages) {
  console.log(`DLQ Reason: ${msg.deadLetterReason}`);
  // Reprocess or log
  await dlqReceiver.completeMessage(msg);
}
```

## Scheduled Messages

```typescript
const sender = client.createSender("my-queue");

// Schedule for future delivery
const scheduledTime = new Date(Date.now() + 60000); // 1 minute from now
const sequenceNumber = await sender.scheduleMessages(
  { body: "Delayed message" },
  scheduledTime
);

// Cancel scheduled message
await sender.cancelScheduledMessages(sequenceNumber);
```

## Message Deferral

```typescript
// Defer message for later
await receiver.deferMessage(message);

// Receive deferred message by sequence number
const deferredMessage = await receiver.receiveDeferredMessages(message.sequenceNumber!);
await receiver.completeMessage(deferredMessage[0]);
```

## Peek Messages (Non-Destructive)

```typescript
const receiver = client.createReceiver("my-queue");

// Peek without removing
const peekedMessages = await receiver.peekMessages(10);
for (const msg of peekedMessages) {
  console.log(`Peeked: ${msg.body}`);
}
```

## Key Types

```typescript
import {
  ServiceBusClient,
  ServiceBusSender,
  ServiceBusReceiver,
  ServiceBusSessionReceiver,
  ServiceBusMessage,
  ServiceBusReceivedMessage,
  ProcessMessageCallback,
  ProcessErrorCallback,
} from "@azure/service-bus";
```

## Receive Modes

```typescript
// Peek-Lock (default) - message locked until completed/abandoned
const receiver = client.createReceiver("my-queue", { receiveMode: "peekLock" });
await receiver.completeMessage(message);   // Remove from queue
await receiver.abandonMessage(message);    // Return to queue
await receiver.deferMessage(message);      // Defer for later
await receiver.deadLetterMessage(message); // Move to DLQ

// Receive-and-Delete - message removed immediately
const receiver = client.createReceiver("my-queue", { receiveMode: "receiveAndDelete" });
```

## Best Practices

1. **Use Microsoft Entra Token Credential** - Use `DefaultAzureCredential` for local development; use `ManagedIdentityCredential` or `WorkloadIdentityCredential` for production
2. **Reuse clients** - Create `ServiceBusClient` once, share across senders/receivers
3. **Close resources** - Always close senders/receivers when done
4. **Handle errors** - Implement `processError` callback for subscription receivers
5. **Use sessions for ordering** - When message order matters within a group
6. **Configure dead-letter** - Always handle DLQ messages
7. **Batch sends** - Use `createMessageBatch()` for multiple messages

## Reference Documentation

For detailed patterns, see:

- [Queues vs Topics Patterns](references/queues-topics.md) - Queue/topic patterns, sessions, receive modes, message settlement
- [Error Handling and Reliability](references/error-handling.md) - ServiceBusError codes, DLQ handling, lock renewal, graceful shutdown

````


### `references/error-handling.md`

````markdown
# Error Handling and Reliability

Comprehensive error handling patterns for @azure/service-bus.

## ServiceBusError

All Service Bus errors extend `ServiceBusError` with a `code` property:

```typescript
import { ServiceBusError } from "@azure/service-bus";

try {
  await sender.sendMessages(message);
} catch (error) {
  if (error instanceof ServiceBusError) {
    console.log(`Code: ${error.code}`);
    console.log(`Message: ${error.message}`);
    console.log(`Retryable: ${error.retryable}`);
    
    handleServiceBusError(error);
  }
}
```

## Error Codes Reference

| Code | Description | Retryable | Action |
|------|-------------|-----------|--------|
| `GeneralError` | Unspecified error | Maybe | Log and investigate |
| `MessagingEntityNotFound` | Queue/topic/subscription doesn't exist | No | Check entity name, create entity |
| `MessageLockLost` | Lock expired before settlement | No | Message will be redelivered |
| `MessageNotFound` | Message no longer available | No | Already processed or expired |
| `MessageSizeExceeded` | Message too large (>256KB standard, >100MB premium) | No | Reduce message size or use claim check |
| `MessagingEntityAlreadyExists` | Entity already exists | No | Use existing entity |
| `MessagingEntityDisabled` | Entity is disabled | No | Enable entity in portal |
| `QuotaExceeded` | Namespace quota exceeded | No | Delete messages, upgrade tier |
| `ServiceBusy` | Service temporarily overloaded | Yes | Retry with backoff |
| `ServiceTimeout` | Operation timed out | Yes | Retry with backoff |
| `ServiceCommunicationProblem` | Network/connection issue | Yes | Retry with backoff |
| `SessionCannotBeLocked` | Session locked by another receiver | Yes | Retry or use different session |
| `SessionLockLost` | Session lock expired | No | Re-accept session |
| `UnauthorizedAccess` | Authentication/authorization failed | No | Check credentials/permissions |

## Error Handling by Code

```typescript
function handleServiceBusError(error: ServiceBusError): void {
  switch (error.code) {
    case "MessagingEntityNotFound":
      console.error(`Entity not found: ${error.message}`);
      // Create entity or fix configuration
      break;
      
    case "MessageLockLost":
      console.warn("Message lock lost - will be redelivered");
      // No action needed, message returns to queue
      break;
      
    case "MessageSizeExceeded":
      console.error("Message too large - use claim check pattern");
      // Store payload in blob, send reference
      break;
      
    case "QuotaExceeded":
      console.error("Quota exceeded - namespace is full");
      // Alert ops team, consider cleanup or upgrade
      break;
      
    case "ServiceBusy":
    case "ServiceTimeout":
    case "ServiceCommunicationProblem":
      if (error.retryable) {
        console.warn(`Transient error: ${error.code} - will retry`);
        // SDK handles retry automatically
      }
      break;
      
    case "SessionCannotBeLocked":
      console.warn("Session busy - trying another session");
      // Use acceptNextSession() instead
      break;
      
    case "SessionLockLost":
      console.warn("Session lock lost - re-accepting session");
      // Re-accept the session
      break;
      
    case "UnauthorizedAccess":
      console.error("Authorization failed - check credentials");
      // Verify RBAC roles or connection string
      break;
      
    default:
      console.error(`Unexpected error: ${error.code} - ${error.message}`);
  }
}
```

## ProcessErrorArgs in Subscribe

When using `receiver.subscribe()`, errors are delivered via `processError`:

```typescript
import { ProcessErrorArgs } from "@azure/service-bus";

receiver.subscribe({
  processMessage: async (message) => {
    // Process message
  },
  processError: async (args: ProcessErrorArgs) => {
    console.error(`Error source: ${args.errorSource}`);
    console.error(`Entity path: ${args.entityPath}`);
    console.error(`Namespace: ${args.fullyQualifiedNamespace}`);
    
    if (args.error instanceof ServiceBusError) {
      console.error(`Code: ${args.error.code}`);
      console.error(`Retryable: ${args.error.retryable}`);
    }
    
    // Error sources:
    // - "receive" - Error receiving messages
    // - "processMessageCallback" - Error in your processMessage handler
    // - "renewLock" - Error renewing message lock
    // - "complete" / "abandon" / "deadLetter" - Settlement errors
    
    switch (args.errorSource) {
      case "receive":
        console.log("Connection issue - SDK will reconnect");
        break;
      case "processMessageCallback":
        console.log("Bug in message handler - fix code");
        break;
      case "renewLock":
        console.log("Lock renewal failed - message may be redelivered");
        break;
    }
  },
});
```

## Dead Letter Queue Handling

Messages that can't be processed go to the dead letter queue (DLQ):

```typescript
// Create DLQ receiver
const dlqReceiver = client.createReceiver("my-queue", {
  subQueueType: "deadLetter",
});

// Process dead letters
const deadLetters = await dlqReceiver.receiveMessages(10);

for (const message of deadLetters) {
  console.log(`Dead letter reason: ${message.deadLetterReason}`);
  console.log(`Error description: ${message.deadLetterErrorDescription}`);
  console.log(`Original body: ${JSON.stringify(message.body)}`);
  console.log(`Delivery count: ${message.deliveryCount}`);
  console.log(`Enqueued time: ${message.enqueuedTimeUtc}`);
  
  // Analyze and fix the issue
  if (canReprocess(message)) {
    // Resend to main queue
    const sender = client.createSender("my-queue");
    await sender.sendMessages({ body: message.body });
    await sender.close();
  }
  
  // Remove from DLQ
  await dlqReceiver.completeMessage(message);
}

await dlqReceiver.close();
```

### DLQ for Topics

```typescript
// DLQ receiver for topic subscription
const dlqReceiver = client.createReceiver(
  "my-topic",
  "my-subscription",
  { subQueueType: "deadLetter" }
);
```

### Automatic Dead Lettering

Messages are automatically dead-lettered when:
- `deliveryCount` exceeds `maxDeliveryCount` (default: 10)
- Message TTL expires (if `deadLetteringOnMessageExpiration` is enabled)
- Subscription filter evaluation fails

## Graceful Shutdown

Always close resources in the correct order:

```typescript
const client = new ServiceBusClient(namespace, credential);
const sender = client.createSender("my-queue");
const receiver = client.createReceiver("my-queue");

// Subscribe to messages
const subscription = receiver.subscribe({
  processMessage: async (msg) => { /* ... */ },
  processError: async (args) => { /* ... */ },
});

// Graceful shutdown handler
async function shutdown(): Promise<void> {
  console.log("Shutting down...");
  
  // 1. Stop receiving new messages
  await subscription.close();
  
  // 2. Close receiver (waits for in-flight messages)
  await receiver.close();
  
  // 3. Close sender (waits for pending sends)
  await sender.close();
  
  // 4. Close client last
  await client.close();
  
  console.log("Shutdown complete");
}

process.on("SIGTERM", shutdown);
process.on("SIGINT", shutdown);
```

## Lock Renewal

### Automatic Lock Renewal (Subscribe)

```typescript
// subscribe() automatically renews locks
receiver.subscribe({
  processMessage: async (message) => {
    // Lock is auto-renewed while processing
    await longRunningOperation(message.body);
  },
  processError: async (args) => { /* ... */ },
}, {
  // Max time to auto-renew (default: 5 minutes)
  maxAutoLockRenewalDurationInMs: 10 * 60 * 1000, // 10 minutes
});
```

### Manual Lock Renewal (receiveMessages)

```typescript
const [message] = await receiver.receiveMessages(1);

// For long processing, manually renew lock
const renewalInterval = setInterval(async () => {
  try {
    await receiver.renewMessageLock(message);
    console.log("Lock renewed");
  } catch (error) {
    console.error("Lock renewal failed:", error);
    clearInterval(renewalInterval);
  }
}, 30000); // Renew every 30 seconds

try {
  await longRunningOperation(message.body);
  await receiver.completeMessage(message);
} finally {
  clearInterval(renewalInterval);
}
```

### Session Lock Renewal

```typescript
const sessionReceiver = await client.acceptSession("my-queue", "session-123");

// Auto-renewal for sessions
const sessionReceiver = await client.acceptSession("my-queue", "session-123", {
  maxAutoLockRenewalDurationInMs: 10 * 60 * 1000,
});

// Manual renewal
await sessionReceiver.renewSessionLock();
```

## Connection Recovery

The SDK automatically handles connection recovery:

```typescript
// SDK reconnects automatically on transient failures
// No manual reconnection code needed

receiver.subscribe({
  processMessage: async (message) => {
    // Processing continues after reconnection
  },
  processError: async (args) => {
    if (args.errorSource === "receive") {
      // Connection issue - SDK is reconnecting
      console.log("Connection lost, SDK reconnecting...");
    }
  },
});
```

## Retry Configuration

Configure retry behavior at client level:

```typescript
import { ServiceBusClient } from "@azure/service-bus";

const client = new ServiceBusClient(namespace, credential, {
  retryOptions: {
    maxRetries: 3,
    retryDelayInMs: 1000,
    maxRetryDelayInMs: 30000,
    mode: "Exponential", // or "Fixed"
  },
});
```

## Idempotent Processing

Design for at-least-once delivery:

```typescript
receiver.subscribe({
  processMessage: async (message) => {
    const messageId = message.messageId;
    
    // Check if already processed (use database, Redis, etc.)
    if (await isAlreadyProcessed(messageId)) {
      console.log(`Duplicate message: ${messageId}`);
      return; // Message will be completed
    }
    
    // Process message
    await processOrder(message.body);
    
    // Mark as processed
    await markAsProcessed(messageId);
  },
  processError: async (args) => { /* ... */ },
});
```

## Poison Message Handling

Handle messages that repeatedly fail:

```typescript
receiver.subscribe({
  processMessage: async (message) => {
    // Check delivery count
    if (message.deliveryCount > 5) {
      console.warn(`Message ${message.messageId} failed ${message.deliveryCount} times`);
      
      // Dead letter with reason
      await receiver.deadLetterMessage(message, {
        deadLetterReason: "MaxRetriesExceeded",
        deadLetterErrorDescription: `Failed after ${message.deliveryCount} attempts`,
      });
      return;
    }
    
    try {
      await processMessage(message.body);
    } catch (error) {
      // Abandon to retry
      await receiver.abandonMessage(message, {
        propertiesToModify: {
          lastError: error.message,
          lastAttempt: new Date().toISOString(),
        },
      });
    }
  },
  processError: async (args) => { /* ... */ },
}, {
  autoCompleteMessages: false, // Manual settlement
});
```

## Best Practices Summary

1. **Always handle errors** - Implement `processError` callback
2. **Use peek-lock mode** - Ensures at-least-once delivery
3. **Design for idempotency** - Messages may be delivered multiple times
4. **Monitor dead letter queues** - Set up alerts for DLQ messages
5. **Configure appropriate lock duration** - Match to processing time
6. **Use auto-lock renewal** - For long-running operations
7. **Graceful shutdown** - Close resources in correct order
8. **Log error codes** - Helps diagnose issues
9. **Set maxDeliveryCount** - Prevent infinite retry loops
10. **Handle poison messages** - Dead letter after max retries

````


### `references/queues-topics.md`

````markdown
# Queues vs Topics Patterns

Detailed patterns for Azure Service Bus queues and topics with @azure/service-bus.

## Queue Patterns (Point-to-Point)

Queues deliver each message to exactly one consumer. Use for work distribution.

### Basic Queue Sender

```typescript
import { ServiceBusClient, ServiceBusMessage } from "@azure/service-bus";
import { DefaultAzureCredential } from "@azure/identity";

const client = new ServiceBusClient(
  process.env.SERVICEBUS_NAMESPACE!,
  new DefaultAzureCredential()
);

const sender = client.createSender("order-queue");

// Send single message
const message: ServiceBusMessage = {
  body: { orderId: "12345", amount: 99.99 },
  contentType: "application/json",
  messageId: "unique-id-12345",
  correlationId: "request-abc",
  applicationProperties: {
    priority: "high",
    source: "web-app",
  },
};

await sender.sendMessages(message);
await sender.close();
```

### Batch Sending (Recommended for Multiple Messages)

```typescript
const sender = client.createSender("order-queue");

// Create batch with size limits
const batch = await sender.createMessageBatch();

const orders = [
  { orderId: "001", amount: 50 },
  { orderId: "002", amount: 75 },
  { orderId: "003", amount: 100 },
];

for (const order of orders) {
  const added = batch.tryAddMessage({ body: order });
  if (!added) {
    // Batch is full - send current batch and create new one
    await sender.sendMessages(batch);
    const newBatch = await sender.createMessageBatch();
    newBatch.tryAddMessage({ body: order });
  }
}

// Send remaining messages
if (batch.count > 0) {
  await sender.sendMessages(batch);
}

await sender.close();
```

### Queue Receiver (Pull Model)

```typescript
const receiver = client.createReceiver("order-queue", {
  receiveMode: "peekLock", // Default - message locked until settled
});

// Receive batch of messages
const messages = await receiver.receiveMessages(10, {
  maxWaitTimeInMs: 5000,
});

for (const message of messages) {
  try {
    const order = message.body;
    console.log(`Processing order: ${order.orderId}`);
    
    // Process successfully - remove from queue
    await receiver.completeMessage(message);
  } catch (error) {
    // Processing failed - return to queue for retry
    await receiver.abandonMessage(message);
  }
}

await receiver.close();
```

### Queue Subscriber (Push Model - Event-Driven)

```typescript
const receiver = client.createReceiver("order-queue");

const subscription = receiver.subscribe({
  processMessage: async (message) => {
    console.log(`Received: ${JSON.stringify(message.body)}`);
    // Message auto-completed on success when using subscribe()
  },
  processError: async (args) => {
    console.error(`Error source: ${args.errorSource}`);
    console.error(`Error: ${args.error.message}`);
    
    if (args.error.code === "MessageLockLost") {
      console.log("Message lock expired - will be redelivered");
    }
  },
}, {
  autoCompleteMessages: true, // Default
  maxConcurrentCalls: 5,
});

// Graceful shutdown
process.on("SIGTERM", async () => {
  await subscription.close();
  await receiver.close();
  await client.close();
});
```

## Topic Patterns (Publish-Subscribe)

Topics deliver messages to multiple subscriptions. Use for fan-out scenarios.

### Topic Publisher

```typescript
// Sending to a topic is identical to sending to a queue
const sender = client.createSender("order-events");

await sender.sendMessages({
  body: { event: "order.created", orderId: "12345" },
  applicationProperties: {
    eventType: "order.created",
    region: "us-west",
  },
  subject: "orders/created", // Can be used for filtering
});

await sender.close();
```

### Subscription Receiver

```typescript
// Receive from a specific subscription
const receiver = client.createReceiver(
  "order-events",      // Topic name
  "billing-service"    // Subscription name
);

const messages = await receiver.receiveMessages(10);

for (const message of messages) {
  console.log(`Billing received: ${message.body.event}`);
  await receiver.completeMessage(message);
}
```

### Multiple Subscriptions (Fan-Out)

```typescript
// Each subscription gets a copy of every message
// Create receivers for different services

const billingReceiver = client.createReceiver("order-events", "billing");
const inventoryReceiver = client.createReceiver("order-events", "inventory");
const notificationReceiver = client.createReceiver("order-events", "notifications");

// Each processes independently
billingReceiver.subscribe({
  processMessage: async (msg) => {
    await processBilling(msg.body);
  },
  processError: async (args) => console.error(args.error),
});

inventoryReceiver.subscribe({
  processMessage: async (msg) => {
    await updateInventory(msg.body);
  },
  processError: async (args) => console.error(args.error),
});
```

## Session-Enabled Queues (Ordered Processing)

Sessions guarantee FIFO ordering within a session and enable stateful processing.

### Send Session Messages

```typescript
const sender = client.createSender("workflow-queue");

// All messages with same sessionId processed in order by same receiver
await sender.sendMessages([
  { body: { step: 1, action: "validate" }, sessionId: "order-123" },
  { body: { step: 2, action: "charge" }, sessionId: "order-123" },
  { body: { step: 3, action: "fulfill" }, sessionId: "order-123" },
]);
```

### Accept Specific Session

```typescript
// Lock a specific session
const sessionReceiver = await client.acceptSession(
  "workflow-queue",
  "order-123"
);

// Process all messages for this session in order
const messages = await sessionReceiver.receiveMessages(10);

for (const message of messages) {
  console.log(`Step ${message.body.step}: ${message.body.action}`);
  await sessionReceiver.completeMessage(message);
}

await sessionReceiver.close();
```

### Accept Next Available Session

```typescript
// Accept any available session (load balancing)
const sessionReceiver = await client.acceptNextSession("workflow-queue", {
  maxAutoLockRenewalDurationInMs: 300000, // 5 minutes
});

console.log(`Processing session: ${sessionReceiver.sessionId}`);

// Get/set session state for checkpointing
const state = await sessionReceiver.getSessionState();
if (state) {
  const checkpoint = JSON.parse(state.toString());
  console.log(`Resuming from step: ${checkpoint.lastStep}`);
}

// Process messages...

// Save checkpoint
await sessionReceiver.setSessionState(
  Buffer.from(JSON.stringify({ lastStep: 3, status: "complete" }))
);

await sessionReceiver.close();
```

## Receive Modes

### Peek-Lock (Default - At-Least-Once)

```typescript
const receiver = client.createReceiver("my-queue", {
  receiveMode: "peekLock",
});

const [message] = await receiver.receiveMessages(1);

// Message is locked - other receivers can't see it
// You MUST settle the message:

await receiver.completeMessage(message);   // Success - remove from queue
await receiver.abandonMessage(message);    // Failure - return to queue
await receiver.deferMessage(message);      // Defer - retrieve by sequence number
await receiver.deadLetterMessage(message); // Move to DLQ
```

### Receive-and-Delete (At-Most-Once)

```typescript
const receiver = client.createReceiver("my-queue", {
  receiveMode: "receiveAndDelete",
});

const [message] = await receiver.receiveMessages(1);

// Message is immediately deleted - no settlement needed
// If processing fails, message is lost
console.log(`Received and deleted: ${message.body}`);
```

## Message Settlement Actions

```typescript
const receiver = client.createReceiver("my-queue");
const [message] = await receiver.receiveMessages(1);

// COMPLETE: Processing succeeded, remove message
await receiver.completeMessage(message);

// ABANDON: Processing failed, return to queue for retry
// Increments deliveryCount
await receiver.abandonMessage(message, {
  propertiesToModify: { retryReason: "timeout" },
});

// DEFER: Can't process now, retrieve later by sequence number
await receiver.deferMessage(message);
// Later: await receiver.receiveDeferredMessages([message.sequenceNumber!]);

// DEAD LETTER: Poison message, move to DLQ
await receiver.deadLetterMessage(message, {
  deadLetterReason: "ValidationFailed",
  deadLetterErrorDescription: "Missing required field: customerId",
});
```

## Scheduled Messages

```typescript
const sender = client.createSender("my-queue");

// Schedule for future delivery
const scheduledTime = new Date(Date.now() + 60 * 60 * 1000); // 1 hour from now

const sequenceNumbers = await sender.scheduleMessages(
  [
    { body: "Reminder: Complete your order" },
    { body: "Your cart is waiting" },
  ],
  scheduledTime
);

console.log(`Scheduled messages: ${sequenceNumbers}`);

// Cancel scheduled messages if needed
await sender.cancelScheduledMessages(sequenceNumbers);
```

## Peek Messages (Non-Destructive)

```typescript
const receiver = client.createReceiver("my-queue");

// Peek without removing or locking
const peekedMessages = await receiver.peekMessages(10);

for (const msg of peekedMessages) {
  console.log(`Peeked: ${msg.body}`);
  console.log(`Sequence: ${msg.sequenceNumber}`);
  console.log(`Enqueued: ${msg.enqueuedTimeUtc}`);
}

// Peek from specific sequence number
const moreMessages = await receiver.peekMessages(10, {
  fromSequenceNumber: 100n,
});
```

## When to Use Queues vs Topics

| Scenario | Use |
|----------|-----|
| Work distribution (one consumer per message) | Queue |
| Fan-out (multiple consumers per message) | Topic |
| Competing consumers (load balancing) | Queue |
| Event broadcasting | Topic |
| Request-reply pattern | Queue with reply-to |
| Ordered processing | Session-enabled Queue |
| Filtered delivery | Topic with subscription filters |

````