Immutable. This exact content is served forever at /api/v1/blob/a7963bf12673c6b2.
--- name: deployment-engineer description: Plan, implement, review, or troubleshoot deployment paths, hosting, CI/CD, release, rollback, environment, and production-readiness decisions for projects. metadata: short-description: Plan simple, safe deployment paths --- # deployment-engineer ## Purpose During Development Use this skill when Codex and similar AI coding agents need to plan, implement, review, or troubleshoot deployment paths for a project. This is a reusable development-time reasoning framework. It is not runtime code and must not be coupled into the product being deployed. ## When To Apply - choosing hosting, runtime, CI/CD, release, rollback, or environment strategy - preparing a project for local preview, staging, production, or demos - reviewing deployment configuration, infrastructure-as-code, containerization, build scripts, or environment variables - reducing deployment complexity while preserving security and operational needs ## What To Inspect - package managers, build scripts, lockfiles, runtime versions, and framework conventions - existing deployment files, CI workflows, Dockerfiles, compose files, IaC, hosting config, and environment examples - secrets handling, environment variables, domains, TLS, storage, database, queue, and background job needs - repository guidance for security, operations, testing, and cost - current production or staging assumptions if the repo already deploys ## Rules To Enforce - choose the simplest deployment target that satisfies the project’s runtime and traffic needs - prefer existing deployment tooling when it is already working and passes security review - keep dev, preview, staging, and production differences explicit - avoid introducing containers, orchestration, IaC, or multi-cloud complexity before the project needs it - require rollback or recovery guidance for production-facing changes - verify build commands and deployment assumptions before presenting them as done - coordinate with `finance-engineer` before adopting paid or usage-based infrastructure ## Common Anti-Patterns To Catch - deploying from undocumented local state - hardcoding secrets, URLs, regions, tenants, or credentials - adding Kubernetes, queues, or managed databases for a static or simple app - missing health checks, migration order, rollback notes, or environment documentation - ignoring cold starts, build limits, storage persistence, or regional constraints ## Expected Output From The Agent - deployment target recommendation - environment and secrets plan - build and release steps - rollback or recovery plan - verification commands - operational handoff notes - cost-impact notes ## Output Structure - Summary - Key Findings - Risks - Affected Files - Affected Areas - Deployment Path - Environment And Secrets - Rollback Plan - Suggested Improvements - Suggested Tests - Skills applied ## Example Usage "Use this skill to deploy this project with the simplest feasible hosting setup." "Review these deployment files for release, environment, rollback, and cost risk."