1---2name: devops-rollout-plan3description: Generate production-ready DevOps rollout plans for infrastructure, application, configuration, and data changes, including preflight checks, phased deployment, verification signals, rollback procedures, communication plan, contingency handling, and post-deployment tasks. Use when the user asks for a rollout plan, deployment plan, release plan, production change plan, or go/no-go checklist.4---56<!-- Generated from harness/github-copilot/skills/devops-rollout-plan/SKILL.md by harness/claude-code/scripts/convert_from_copilot.py. Edit the source, not this file. -->78# DevOps rollout plan910Create a comprehensive step-by-step rollout plan that turns a proposed infrastructure or application change into a preflighted, observable, reversible production deployment with communication, rollback, and contingency paths.1112## When to invoke1314- "Generate a rollout plan for this deployment."15- "Create a production change plan."16- "Write preflight checks, rollout steps, verification, and rollback."17- "Plan this Kubernetes, VM, serverless, or container release."18- "Prepare a go/no-go checklist for an infrastructure change."1920## Inputs2122Use `$ARGUMENTS` as the change description, environment, or target system when provided. If details are missing, collect enough to avoid a generic plan.2324| Input group | Required details |25| --- | --- |26| Change Description | What's changing: infrastructure, application, configuration; version or state transition from/to; problem solved or feature added. |27| Environment Details | Target environment: dev, staging, production, all; infrastructure type: Kubernetes, VMs, serverless, containers; affected services and dependencies; current capacity and scale. |28| Constraints & Requirements | Acceptable downtime window, change window restrictions, approval requirements, regulatory or compliance considerations. |29| Risk Assessment | Blast radius, data migrations or schema changes, rollback complexity and safety, known risks. |3031## Rollout structure3233| Section | Required content |34| --- | --- |35| Executive Summary | What, why, when, duration, risk level, rollback time, affected systems, user impact, expected downtime. |36| Prerequisites & Approvals | Technical lead, security, compliance, business approvals; resources, capacity, backups, monitoring, rollback automation; pre-deployment backups. |37| Preflight Checks | Infrastructure health, application health baseline, dependencies, monitoring baseline metrics, go/no-go checklist. |38| Step-by-Step Rollout Procedure | Phases for pre-deployment, deployment, progressive verification; specific commands, validation after each step, duration estimates. |39| Verification Signals | Immediate, short-term, medium-term, and long-term checks. |40| Rollback Procedure | Decision criteria, automated or manual rollback steps, infrastructure revert or full restore, post-rollback verification, stakeholder notification. |41| Communication Plan | T-24h notice, deployment start, periodic progress updates, completion, rollback notice. |42| Post-Deployment Tasks | Immediate 1h, short-term 24h, medium-term 1 week review and lessons learned. |43| Contingency Plans | Partial failure, performance degradation, data inconsistency, dependency failure. |44| Contact Information | Primary and secondary on-call, escalation path, emergency contacts for infrastructure, security, database, networking. |4546## Verification windows4748| Window | Time | Signals |49| --- | --- | --- |50| Immediate | 0-2 min | Deployment success, pods/containers started, health checks passing. |51| Short-term | 2-5 min | Application responding, error rates acceptable, latency normal. |52| Medium-term | 5-15 min | Sustained metrics, stable connections, integrations working. |53| Long-term | 15+ min | No degradation, capacity healthy, business metrics normal. |5455## Customization rules5657| Dimension | Adaptation |58| --- | --- |59| Infrastructure Type | Kubernetes plans include pods, rollout status, readiness/liveness probes, node capacity; VMs include image/version, service restart, health endpoint; serverless includes function versions, aliases, cold start, trigger health; databases include backups, migrations, locks, replication lag. |60| Risk Level | Low risk gets simplified gates; medium risk gets standard preflight, canary, rollback; high risk gets additional approvals, dry run, checkpoint, and explicit abort criteria. |61| Change Type | Code deployment, infrastructure, configuration, and data migration need different validation and rollback mechanics. |62| Environment | Production uses the full plan; staging can be simplified; development is minimal but still verifies success. |6364## Rollout principles6566- Always have a tested rollback plan.67- Communicate early and often.68- Monitor metrics, not just logs.69- Document everything.70- Learn from each deployment.71- Never deploy on Friday afternoon unless the change is critical.72- Never skip verification steps.73- Never assume "it should work".7475## Output template7677```markdown78# Rollout Plan: <change name>7980## 1. Executive Summary81- What: <change>82- Why: <reason>83- When: <window>84- Duration: <estimate>85- Risk level: low | medium | high86- Rollback time: <estimate>87- Affected systems: <systems>88- User impact / expected downtime: <impact>8990## 2. Prerequisites & Approvals91| Requirement | Owner | Status |92| --- | --- | --- |93| Technical lead approval | <name> | pending |9495## 3. Preflight Checks96- [ ] Infrastructure health baseline captured.97- [ ] Application health baseline captured.98- [ ] Dependencies available.99- [ ] Monitoring baseline metrics captured.100- [ ] Go/no-go decision recorded.101102## 4. Step-by-Step Rollout Procedure103| Phase | Step | Command/action | Validation | Duration |104| --- | --- | --- | --- | --- |105| Pre-deployment | <step> | `<command>` | <signal> | <time> |106107## 5. Verification Signals108| Window | Signals | Pass criteria |109| --- | --- | --- |110| Immediate (0-2 min) | deployment, pods/containers, health checks | <criteria> |111| Short-term (2-5 min) | response, error rate, latency | <criteria> |112| Medium-term (5-15 min) | sustained metrics, connections, integrations | <criteria> |113| Long-term (15+ min) | degradation, capacity, business metrics | <criteria> |114115## 6. Rollback Procedure116**Decision criteria:** <when to rollback>117**Steps:** <automated/infrastructure/full restore steps>118**Post-rollback verification:** <checks>119**Communication:** <notification>120121## 7. Communication Plan122| Time | Audience | Channel | Message |123| --- | --- | --- | --- |124| T-24h | <stakeholders> | <channel> | <notice> |125126## 8. Post-Deployment Tasks127- Immediate (1h): <tasks>128- Short-term (24h): <tasks>129- Medium-term (1 week): <tasks>130131## 9. Contingency Plans132| Scenario | Symptoms | Response | Timeline |133| --- | --- | --- | --- |134| Partial failure | <symptoms> | <response> | <time> |135136## 10. Contact Information137| Role | Primary | Secondary | Escalation |138| --- | --- | --- | --- |139| On-call | <person> | <person> | <path> |140```141142## Quality gate143144- [ ] Change description, environment, constraints, and risk assessment are captured.145- [ ] The plan includes all ten sections from Executive Summary through Contact Information.146- [ ] Preflight checks include infrastructure, application, dependencies, monitoring, and go/no-go.147- [ ] Rollout steps include commands/actions, validation after each step, and duration estimates.148- [ ] Verification signals cover 0-2 min, 2-5 min, 5-15 min, and 15+ min windows.149- [ ] Rollback has decision criteria, steps, post-rollback verification, and communication.150- [ ] Communication includes T-24h, deployment start, progress updates, completion, and rollback notices.151- [ ] Contingencies cover partial failure, performance degradation, data inconsistency, and dependency failure.