AI Prompts for Claude for DevOps

Write a GitHub Actions workflow for [DESCRIBE STACK] that runs lint, type check, tests, and deploys to [DESCRIBE TARGET] on merge to main. Add caching to make it fast.

My pipeline takes [N] minutes and it is too slow: [PASTE OR DESCRIBE WORKFLOW]. Add caching, parallelise independent jobs, and tell me where the biggest time savings are.

Write a deployment pipeline that deploys to staging automatically and requires manual approval before production. Include Slack notifications for each stage.

I need a Docker build pipeline that uses multi-stage builds for [DESCRIBE APP], caches layers effectively, and pushes to [DESCRIBE REGISTRY].

My CI pipeline fails intermittently: [DESCRIBE FLAKY TEST OR STEP]. Diagnose the most likely causes and show how to fix or retry the flaky step properly.

Write a Terraform configuration for [DESCRIBE INFRASTRUCTURE, E.G. "ECS CLUSTER WITH ALB AND RDS"]. Use modules, separate environment-specific variables, and add outputs for key resource IDs.

Write a Docker Compose file for local development of [DESCRIBE APP WITH ITS DEPENDENCIES]. Include all services with correct port mappings, environment variables, and volume mounts.

Write a Kubernetes manifest for [DESCRIBE SERVICE] with Deployment, Service, and HorizontalPodAutoscaler. Set resource requests and limits appropriate for [DESCRIBE LOAD].

I need to provision [DESCRIBE AWS/GCP/AZURE RESOURCE] with Terraform. Show the resource definitions, the IAM permissions needed, and the variables to make it reusable across environments.

My Terraform state is a mess. Explain how to refactor it into modules without losing existing resources, including how to move resources between state files safely.

Write a bash script that [DESCRIBE TASK]. Include error handling with set -e and set -o pipefail, meaningful error messages, and a --dry-run flag.

I need to automate [DESCRIBE DEPLOYMENT TASK, E.G. "DATABASE MIGRATIONS IN THE DEPLOY PIPELINE"]. Write the script with a timeout, rollback on failure, and a notification on completion.

Write a script to clean up [DESCRIBE STALE CLOUD RESOURCES, E.G. "UNUSED ECR IMAGES OLDER THAN 30 DAYS"]. Preview what will be deleted before doing anything destructive.

How do I securely manage secrets in [DESCRIBE ENVIRONMENT]? Show me the approach using [AWS SECRETS MANAGER / VAULT / GITHUB SECRETS] and how to inject them into the application.

Write a monitoring script that checks [DESCRIBE HEALTH CHECKS] every [N MINUTES] and pages [SLACK / PAGERDUTY] if any check fails.

My deployment just failed with this error: [PASTE ERROR]. Walk me through diagnosing the root cause step by step and suggest the most likely fixes in order.

My service has [DESCRIBE SYMPTOM: HIGH LATENCY / INCREASED ERROR RATE / MEMORY LEAK]. Here are the relevant metrics: [DESCRIBE]. What are the most likely causes and how do I narrow it down?

Write the incident post-mortem template for [DESCRIBE INCIDENT]. Include: timeline, root cause, contributing factors, business impact, immediate fix, and long-term prevention tasks.

I need to roll back this deployment quickly: [DESCRIBE DEPLOYMENT]. Write the exact commands to execute the rollback, verify it, and communicate status to stakeholders.

Set up distributed tracing for this [DESCRIBE ARCHITECTURE]. Recommend the tooling, show how to instrument the services, and explain how to trace a slow request end-to-end.