Cloud migration planning -- 7R assessment, workload classification, wave planning, cutover. Use when the user asks to "plan cloud migration", "assess workloads for migration", "design landing zone", "create migration waves", "plan cutover strategy", or mentions 7R, rehost, replatform, refactor, lift-and-shift, or migration factory.
From pmnpx claudepluginhub javimontano/mao-pm-apexThis skill is limited to using the following tools:
examples/README.mdexamples/sample-output.htmlexamples/sample-output.mdprompts/metaprompts.mdprompts/use-case-prompts.mdreferences/body-of-knowledge.mdreferences/knowledge-graph.mmdreferences/migration-patterns.mdreferences/state-of-the-art.mdSearches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.
Estimates prompt tokens and complexity to offer Brief, Standard, Detailed, or Exhaustive response options with projected token budgets before answering.
Cloud migration moves workloads from on-premises or legacy environments to cloud platforms. This skill produces comprehensive migration plans covering 7R assessment, workload analysis, wave planning, landing zone design, execution patterns, and post-migration optimization.
Migrar sin estrategia 7R es mover problemas de datacenter a la nube. Cada workload merece una clasificación explícita (rehost, replatform, refactor, repurchase, retire, retain, relocate). Wave planning reduce riesgo. Cutover rehearsal es obligatorio — nunca se hace un cutover en producción sin haber ensayado en staging.
The user provides a migration program or portfolio name as $ARGUMENTS. Parse $1 as the program/portfolio name used throughout all output artifacts.
Parameters:
{MODO}: piloto-auto (default) | desatendido | supervisado | paso-a-paso
{FORMATO}: markdown (default) | html | dual{VARIANTE}: ejecutiva (~40% — S1 7R classification + S3 wave plan + S5 cutover) | técnica (full 6 sections, default)Before generating migration plan, detect existing infrastructure context:
!find . -name "*.tf" -o -name "*.yaml" -o -name "inventory*" -o -name "*.csv" | head -20
If reference materials exist, load them:
Read ${CLAUDE_SKILL_DIR}/references/migration-patterns.md
Classify every workload using the 7R framework to determine optimal migration strategy.
7R Strategies:
Classification Decision Matrix:
| Factor | Rehost | Replatform | Refactor |
|---|---|---|---|
| Business criticality | Low-medium | Medium | High (strategic) |
| Technical complexity | Any | Low-medium | High (code access required) |
| Timeline pressure | Urgent (<3 months) | Moderate (3-6 months) | Long-term (6-18 months) |
| Cloud benefit needed | Minimal | Moderate | Maximum |
| Budget | Low ($) | Medium ($$) | High ($$$) |
Output: Application-by-application classification table with strategy, rationale, effort estimate, risk level, and wave assignment.
Automated Discovery Tools:
| Tool | Provider | Method | Discovers |
|---|---|---|---|
| AWS Application Discovery Service | AWS | Agent or agentless | Server config, utilization, network connections, process data |
| AWS Migration Hub | AWS | Aggregator | Unified view across discovery + migration tools |
| Azure Migrate | Azure | Agent or appliance | Servers, databases, web apps, VDI; includes assessment + migration |
| Azure App Service Migration Assistant | Azure | Web app scan | .NET/Java web app readiness for Azure App Service |
| Google Cloud Migration Center | GCP | Agent or API import | Technical fit, TCO, migration complexity scoring |
| Cloudamize | Multi-cloud | Agent-based | Performance profiling, right-sizing, cost projection, dependency mapping |
| Flexera One | Multi-cloud | CMDB integration | License optimization, cloud cost modeling, portfolio rationalization |
Application Inventory Fields: Name, owner, business unit, criticality tier (T1-T4), technology stack, infrastructure requirements (CPU, memory, storage, network), performance baseline (response time, throughput, utilization), compliance requirements.
Dependency Graph:
Data Gravity Analysis:
(data_volume_TB * access_frequency * compliance_anchors) -- higher score = harder to move(data_size_GB / bandwidth_Gbps) * 8 / 3600 = hours + 30-50% verification overheadApplication Difficulty Scoring:
(technical * 0.3) + (dependencies * 0.25) + (data_gravity * 0.25) + (compliance * 0.2)Migration Factory Model:
A migration factory standardizes repeatable processes across waves. Structure:
| Role | Responsibility | Staffing |
|---|---|---|
| Factory Manager | Wave scheduling, escalation, metrics | 1 per program |
| Migration Architect | Technical design per application | 1 per 10-15 apps |
| Migration Engineer | Execute rehost/replatform using tooling | 2-4 per wave |
| Test Lead | Validation scripts, acceptance criteria | 1 per wave |
| App Owner (from business) | Requirements, UAT sign-off, cutover approval | 1 per application |
Wave Design:
Pilot Selection Criteria:
Wave Calendar:
Account Structure:
Networking:
Security Baseline:
Governance Guardrails:
Rehost Tools: AWS MGN (Application Migration Service), Azure Migrate/Site Recovery, GCP Migrate for Compute Engine. Process: install agent, replicate, test, cutover, decommission.
Database Migration: Homogeneous (native replication, DMS continuous) or heterogeneous (schema conversion + DMS). Zero-downtime via continuous replication. Validate: row counts, checksums, sample comparison.
TCO Calculator Methodology:
| Cost Category | On-Premises | Cloud |
|---|---|---|
| Compute | Hardware amortization (3-5yr) | Reserved + on-demand instances |
| Storage | SAN/NAS, disk replacement | S3/EBS tiered pricing |
| Facilities | Power, cooling, DC lease ($8-15/kW/mo) | Included |
| Staff | FTE * fully loaded cost | Reduced (managed services) |
| Licenses | On-prem perpetual or subscription | BYOL, cloud-native, or included |
| Network | WAN circuits, load balancers | Egress fees, cross-AZ traffic |
| Migration (one-time) | N/A | Dual-run, tooling, consulting, training |
Cutover Rehearsal Checklist:
Perform a full dress rehearsal 1-2 weeks before each production cutover:
Rollback Decision Criteria:
Trigger rollback if ANY of the following occur during cutover:
| Condition | Threshold | Action |
|---|---|---|
| Error rate spike | >5x baseline for 15+ minutes | Immediate rollback |
| Latency degradation | P99 >3x baseline for 15+ minutes | Immediate rollback |
| Data integrity failure | Any checksum mismatch on critical tables | Immediate rollback |
| Functionality gap | Any Tier 1 feature non-functional | Immediate rollback |
| Cutover window exceeded | >80% of planned window with steps remaining | Assess and likely rollback |
| Monitoring blind spot | Key dashboards/alerts not functioning | Pause and remediate; rollback if >30 min |
Common Anti-Patterns:
| Anti-Pattern | Consequence | Mitigation |
|---|---|---|
| Skip discovery; migrate by guesswork | Missed dependencies cause outages during cutover | Run automated discovery 30+ days |
| Migrate without application owner | No UAT, no sign-off, unclear accountability | Require named owner before wave assignment |
| No cutover rehearsal | Surprises on go-live night, panic rollbacks | Mandatory dress rehearsal for every wave |
| Big-bang migration of entire portfolio | Maximum blast radius, no learning curve | Wave-based with pilot first |
| Underestimate data transfer time | Cutover window exceeded, forced rollback | Calculate transfer time + 50% buffer; use continuous replication |
| Retire nothing | Cloud costs add to existing costs | Enforce retire/retain classification in 7R assessment |
| No rollback plan | Cannot revert when issues found | Define rollback triggers and test procedure |
Parallel-Run Validation:
Functional Validation: Automated test suites, manual smoke testing of critical flows, integration testing with dependent systems, data validation.
Performance Baseline: Compare cloud vs. on-premises for latency, throughput, resource utilization. Document regressions.
Cost Optimization (post-migration):
Decommission:
| Decision | Enables | Constrains | When to Use |
|---|---|---|---|
| Rehost | Speed, low risk | No modernization | Datacenter exit deadline |
| Replatform | Some cloud benefit, moderate effort | Partial optimization | Managed DB, container runtime swaps |
| Refactor | Full cloud-native benefit | High effort, long timeline | Strategic, 5+ year lifecycle apps |
| Migration Factory | Repeatable, metrics-driven, scalable | Setup overhead, process rigidity | >20 applications, enterprise programs |
| Big Bang | Clean cutover, no hybrid period | High risk, long outage | Small portfolios (<10 apps), maintenance windows |
| Wave-Based | Incremental risk, learning curve | Hybrid period, dual costs | Large portfolios, enterprise migrations |
| Direct Connect | High bandwidth, low latency | Cost, 4-8 week lead time | Large data (>5TB), long-term hybrid |
Datacenter Exit with Hard Deadline: Favor rehost for speed. Accept technical debt. Plan post-migration optimization waves. Prioritize by lease expiry.
Shared Database Serving Multiple Applications: Cannot migrate database independently. Options: migrate all consumers together, introduce API layer to decouple, or replicate and gradually cut over consumers.
Mainframe Workloads: Specialized tools (Micro Focus, AWS Mainframe Modernization, Azure Mainframe Migration). Replatform to cloud-hosted emulation first, then gradually refactor.
Compliance-Restricted Data: Data residency may limit regions. Encryption requirements affect replication tooling. Audit trail must be maintained through migration.
No Source Code Available: Rehost is the only viable strategy. Containerization may be possible for binary apps. Evaluate repurchase with SaaS alternative.
Before finalizing delivery, verify:
| Format | Default | Description |
|---|---|---|
markdown | Yes | Rich Markdown + Mermaid diagrams. Token-efficient. |
html | On demand | Branded HTML (Design System). Visual impact. |
dual | On demand | Both formats. |
Default output is Markdown with embedded Mermaid diagrams. HTML generation requires explicit {FORMATO}=html parameter.
Primary: A-01_Cloud_Migration_Plan.html -- Executive summary, 7R classification table, dependency map, wave plan, landing zone design, execution playbook, validation checklist, cost optimization targets.
Secondary: Application inventory spreadsheet, wave calendar, cutover runbook, rollback procedures, TCO comparison, decommission checklist.
| Caso | Estrategia de Manejo |
|---|---|
| Datacenter exit con hard deadline | Favor rehost para velocidad. Aceptar tech debt. Plan post-migration optimization waves. Priorizar por lease expiry. |
| Shared database serving multiple apps | No migrar DB independientemente. Opciones: migrar todos los consumers juntos, API layer para desacoplar, o replicar y gradual cutover. |
| Mainframe workloads | Tools especializados (Micro Focus, AWS Mainframe Modernization). Replatform a cloud-hosted emulation primero, refactor gradual. |
| Compliance-restricted data | Data residency puede limitar regiones. Encryption requirements afectan tooling de replicacion. Audit trail durante migracion. |
| No source code available | Rehost es la unica estrategia viable. Containerizacion posible para binary apps. Evaluar repurchase con SaaS alternative. |
| >200 workloads sin inventario | Deploy agentless discovery minimo 30 dias antes de clasificar. Migration factory model es mandatorio para throughput. |
| Decision | Alternativa Descartada | Justificacion |
|---|---|---|
| 7R framework para clasificacion | Binary migrate/no-migrate, 3R simplificado | 7R (Rehost/Replatform/Refactor/Repurchase/Retire/Retain/Relocate) cubre el espectro completo de opciones. Retire y Retain son decisiones legitimas que evitan costo innecesario. |
| Wave-based migration sobre big-bang | Big-bang de todo el portfolio | Waves reducen blast radius, permiten aprendizaje incremental, y escalan throughput progresivamente. Big-bang maximiza riesgo. |
| Cutover rehearsal mandatorio | Cutover directo en produccion | Rehearsal valida runbook end-to-end, mide duracion real, y prueba rollback. Sin rehearsal, sorpresas en go-live night. |
| Migration factory model para >20 apps | Migracion artesanal por aplicacion | Factory estandariza procesos repetibles, habilita metricas (apps/week), y escala throughput a 15-25 apps/month. |
graph TD
subgraph Core["Conceptos Core"]
SEVR["7R Classification"]
WORKLOAD["Workload Analysis"]
WAVE["Wave Planning"]
LANDING["Landing Zone"]
CUTOVER["Migration Execution"]
VALID["Validation & Optimization"]
end
subgraph Inputs["Entradas"]
INFRA["Infrastructure Inventory"]
APPS["Application Portfolio"]
DEPS["Dependency Map"]
CLOUD["Target Cloud Platform"]
end
subgraph Outputs["Salidas"]
PLAN["Cloud Migration Plan"]
CLASS["7R Classification Table"]
CALENDAR["Wave Calendar"]
TCO["TCO Comparison"]
RUNBOOK["Cutover Runbook"]
end
subgraph Related["Skills Relacionados"]
CLOUDNAT["cloud-native-architecture"]
INFRAARCH["infrastructure-architecture"]
ASIS["asis-analysis (Cloud)"]
ENTARCH["enterprise-architecture"]
end
INFRA --> WORKLOAD
APPS --> SEVR
DEPS --> WAVE
CLOUD --> LANDING
SEVR --> WAVE
WORKLOAD --> SEVR
WAVE --> CUTOVER
LANDING --> CUTOVER
CUTOVER --> VALID
PLAN --> CLASS
PLAN --> CALENDAR
PLAN --> TCO
PLAN --> RUNBOOK
SEVR --> PLAN
CLOUDNAT -.-> SEVR
INFRAARCH -.-> LANDING
ASIS -.-> WORKLOAD
ENTARCH -.-> SEVR
Formato Markdown (default):
# Cloud Migration Plan: {program}
## S1: Migration Assessment & 7R Classification
| Application | Strategy | Rationale | Effort | Risk | Wave |
...
## S2: Workload Analysis & Dependency Mapping
### Dependency Graph (Mermaid)
### Data Gravity Analysis
## S3: Wave Planning & Sequencing
### Wave 0: Foundation (4-6 weeks)
### Wave 1: Pilot (2-4 weeks)
### Wave 2-N: Production
## S4: Landing Zone Design
## S5: Migration Execution Patterns
### Cutover Rehearsal Checklist
### Rollback Decision Criteria
## S6: Validation & Optimization
### TCO Comparison
| Category | On-Prem | Cloud |
...
Formato XLSX (bajo demanda):
Sheet 1: Application Inventory — name, owner, criticality, stack, 7R, wave, effort, risk
Sheet 2: Dependency Matrix — source app, target app, dependency type, coupling level
Sheet 3: Wave Calendar — wave #, apps, dates, duration, go/no-go status
Sheet 4: TCO Comparison — cost categories, on-prem vs cloud, break-even
Sheet 5: Cutover Runbook — step, responsible, duration, rollback, status
Sheet 6: Post-Migration Optimization — right-sizing, RI/SP, storage tiering
Formato HTML (bajo demanda):
A-01_Cloud_Migration_Plan_{cliente}_{WIP}.htmlFormato DOCX (bajo demanda):
{fase}_Cloud_Migration_Plan_{cliente}_{WIP}.docxFormato PPTX (bajo demanda):
{fase}_{entregable}_{cliente}_{WIP}.pptx| Dimension | Peso | Criterio |
|---|---|---|
| Trigger Accuracy | 10% | Activacion correcta ante keywords de cloud migration, 7R, wave planning, landing zone, cutover, lift-and-shift. |
| Completeness | 25% | 6 secciones cubren assessment, workloads, waves, landing zone, execution, y validation. Cada app tiene 7R classification. |
| Clarity | 20% | 7R classification con rationale por aplicacion. Wave sequencing justificado por dependencias. Rollback criteria con thresholds especificos. |
| Robustness | 20% | Edge cases (deadline, shared DB, mainframe, compliance, no source code, >200 workloads) manejados. Anti-patterns documentados. |
| Efficiency | 10% | Variante ejecutiva reduce a S1+S3+S5 (~40%). Migration factory model escala throughput. |
| Value Density | 15% | TCO comparison con break-even. Cutover rehearsal checklist accionable. Rollback decision criteria con thresholds automatizables. |
Umbral minimo: 7/10. Debajo de este umbral, revisar 7R classification completeness y cutover rehearsal coverage.
Autor: Javier Montano · Comunidad MetodologIA | Ultima actualizacion: 15 de marzo de 2026