菜单
Deploy the Your Project Rails application to staging or production. Supports local builds via bin/deploy-staging and CI/CD via branch pushes.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Check application logs from local Docker Compose or remote AWS CloudWatch environments. Supports local, staging, and production targets.
Manage the local Docker Compose development environment. Supports start, stop, restart, and status operations.
Verify that Solid Queue background jobs are running properly. Checks worker health, recurring job execution, and optionally triggers a test job to prove end-to-end processing. Supports local, staging, and production targets.
Verify that OpenTelemetry traces are being collected and exported to X-Ray. Check trace health, find slow requests, investigate errors, and view service dependencies.
Best practices for Rails Action Mailer. Use when writing new mailers, refactoring existing mailers, or when a mailer has inline business logic, missing previews, synchronous delivery, or mixed responsibilities. Applies patterns - single-responsibility mailers, parameterized mailers, deliver_later by default, mailer concerns, service delegation, previews, and structured testing.
Best practices for Rails Active Job with Solid Queue. Use when writing new background jobs, refactoring existing jobs, or when a job has mixed responsibilities, inline business logic, non-idempotent design, or missing error handling. Applies patterns - single-responsibility jobs, argument serialization, idempotent design, retry/discard strategies, queue management, recurring schedules, job concerns, and service delegation.
| name | ops-deploy |
| description | Deploy the Your Project Rails application to staging or production. Supports local builds via bin/deploy-staging and CI/CD via branch pushes. |
| allowed-tools | ["Bash","Read"] |
Deploy the Your Project Rails application to the specified environment.
Target environment: $ARGUMENTS (expected: staging or production)
stagingChoose deployment method:
bin/deploy-stagingstaging branchFor local deploy:
aws sts get-caller-identity --profile your-project-stagingaws sso login --profile your-project-stagingbin/deploy-staging --profile your-project-staging
--service web|worker (default: all), --no-deploy (build/push only), --dry-run (preview)Post-deploy verification:
aws ecs describe-services \
--cluster webCluster \
--services web-rails-service worker-service \
--profile your-project-staging \
--region us-east-1 \
--query 'services[].{name:serviceName,running:runningCount,desired:desiredCount,rollout:deployments[0].rolloutState}' \
--output table
Check logs for errors after services stabilize:
aws logs describe-log-groups \
--profile your-project-staging \
--region us-east-1 \
--query 'logGroups[].logGroupName' \
--output table
Then tail the relevant log group for recent errors.
productionRequire explicit confirmation from the user before proceeding. Do NOT deploy to production without the user confirming they want to deploy.
Important: The production auto-deploy trigger (main branch) is currently commented out in .github/workflows/deploy.yml. Only the staging trigger is active.
CI/CD deploy: merge/push to main branch (if the trigger is re-enabled)
Post-deploy verification (same as staging but with --profile your-project-production)
Deploy now to the specified environment.