Agent Skills
› cosmicstack-labs/mercury-agent-skills
› message-queues
message-queues
GitHub指导设计可靠的消息驱动系统，涵盖RabbitMQ、Kafka、SQS等队列类型及模式。详解竞争消费者、发布订阅、分区消费及死信队列机制，提供代码示例与幂等性、监控等最佳实践。
Trigger Scenarios
消息队列选型与设计
RabbitMQ或Kafka使用问题
死信队列处理方案
分布式系统解耦咨询
Install
npx skills add cosmicstack-labs/mercury-agent-skills --skill message-queues -g -y
SKILL.md
Frontmatter
{
"name": "message-queues",
"metadata": {
"tags": [
"message-queues",
"kafka",
"rabbitmq",
"sqs",
"event-streaming"
],
"author": "cosmicstack-labs",
"version": "1.0.0",
"category": "backend"
},
"description": "RabbitMQ, Kafka, SQS, pub\/sub, competing consumers, dead letter queues, and event streaming"
}
Message Queues
Design reliable message-driven systems.
Queue Types
| Queue | Persistence | Ordering | Use Case |
|---|---|---|---|
| RabbitMQ | Optional | Per queue | Task distribution, RPC |
| Apache Kafka | Durable (disk) | Per partition | Event streaming, logs |
| AWS SQS | Durable | Best effort (std) / Strict (FIFO) | Serverless decoupling |
| Redis Pub/Sub | None | Per channel | Real-time notifications |
RabbitMQ Patterns
Work Queues (Competing Consumers)
Producer → Queue → Consumer 1
→ Consumer 2
→ Consumer 3
- Messages distributed round-robin
- Ack on success, nack on failure (requeue or DLQ)
- Prefetch count controls concurrency
Pub/Sub (Exchange → Binding → Queue)
- Fanout: broadcast to all queues
- Direct: route by routing key
- Topic: route by pattern (user.*, user.created)
- Headers: route by header values
Kafka Patterns
Topics & Partitions
- Messages within a partition are ordered
- Partitions enable parallelism
- Consumer group = one instance per partition
Producer
await producer.send({
topic: 'order-events',
messages: [{ key: orderId, value: JSON.stringify(order) }],
});
Consumer
await consumer.run({
eachMessage: async ({ topic, partition, message }) => {
await processOrder(message.value);
},
});
Dead Letter Queues
- Messages that can't be processed go to DLQ
- Analyze DLQ periodically for systemic issues
- DLQ messages can be replayed after fix
- Set max retry count before DLQ
Best Practices
- Idempotent consumers (same message processed twice = safe)
- Monitor queue depth, consumer lag, error rate
- Set message TTL to prevent infinite backlog
- Use structured message schemas (Avro, Protobuf)
- Test with network failures and consumer crashes
Version History
- 38e2523 Current 2026-07-05 19:36
