Immutable. This exact content is served forever at /api/v1/blob/c8c68fdc5846b2eb.
# operations-engineer ## Purpose During Development Use this skill when Codex and similar AI coding agents need to plan, implement, or review the operational readiness of a project. This is a reusable development-time reasoning framework. It is not runtime code and must not be coupled into the product being operated. ## When To Apply - preparing a project for production, demos, handoff, support, or repeated use - defining logs, metrics, alerts, backups, incident response, runbooks, or maintenance tasks - reviewing reliability, recoverability, support burden, or operational ownership - deciding how much operations work is enough for the simplest feasible launch ## What To Inspect - deployment path, runtime model, data stores, background jobs, queues, storage, and external services - existing observability, logging, monitoring, error reporting, backup, and support tools - failure modes for startup, deploy, rollback, migration, secrets, data loss, rate limits, and dependencies - ownership notes, runbooks, incident docs, support workflows, and on-call expectations - cost and security constraints that affect operations ## Rules To Enforce - define the minimum operational controls needed for the project’s actual risk - prefer existing observability and support tools when they pass security review - make failure modes and recovery actions explicit before launch - keep runbooks short enough to use during stress - avoid expensive monitoring, tracing, or incident systems before the project needs them - coordinate with `deployment-engineer`, `security-engineer`, and `finance-engineer` for production-facing decisions ## Common Anti-Patterns To Catch - launching without knowing how to detect failure - logs that omit useful context or leak sensitive data - backups that are configured but never restored in testing - alerts that page on symptoms nobody can act on - operational plans that require paid tooling for a low-risk prototype - undocumented manual steps that only work on one machine ## Expected Output From The Agent - operational readiness summary - observability plan - failure modes and recovery plan - runbook expectations - backup and data-retention notes - support and maintenance ownership - verification plan ## Output Structure - Summary - Key Findings - Risks - Affected Files - Affected Areas - Operational Readiness - Observability Plan - Recovery Plan - Suggested Improvements - Suggested Tests - Skills applied ## Example Usage "Use this skill to prepare this deployed project for basic production operations." "Review this launch plan for logging, monitoring, backups, runbooks, and support risk."