A Claude skill (with a matching GitHub Copilot prompt-file edition) for describing, designing, comparing, scoring, and recommending cloud operating models in a corporate environment — and producing a phased transition roadmap to move from the current model to a target one.
The flagship output is a layered assessment document: an executive summary any leader can read, backed by the technical depth an architect or platform lead needs to act on, ending in a phased transition roadmap. Optionally, it generates a one-page visual operating-model diagram.
- Describe an operating model in the organization's own context.
- Compare 2–4 candidate models against weighted criteria.
- Recommend a target model with rationale, trade-offs, and preconditions.
- Plan the transition as phases, with RACI/ownership shifts and exit criteria.
- Right-size the team — benchmark required FTE against peers (top-down ratios + bottom-up model).
- Visualize the target model as an adaptable draw.io diagram.
It anchors on vendor-neutral archetypes — Centralized, Decentralized, Federated/Hub-and-Spoke (CCoE), and Platform Engineering — with a cross-walk to Azure CAF, AWS, and Gartner vocabulary.
cloud-operating-model/
├── SKILL.md # workflow, mode-routing, grounding guidance
├── references/
│ ├── operating-models.md # archetype catalog + framework cross-walk
│ ├── evaluation-criteria.md # weighted criteria + scoring rubric
│ ├── transition-roadmap.md # phasing, RACI evolution, failure modes
│ ├── workforce-sizing.md # FTE benchmarking & right-sizing vs. peers
│ └── platform-as-a-product.md # worked Platform Engineering example + techniques
└── assets/
├── assessment-template.md # layered exec + technical output document
├── opmodel-diagram-template.drawio # adaptable three-layer diagram skeleton
└── platform-opmodel-example.drawio # a complete worked diagram (gold standard)
Place the cloud-operating-model/ folder where your Claude skills are discovered, then ask in
natural language — e.g. "help me pick the right cloud operating model and a roadmap to get there"
or "compare a centralized cloud team vs a federated CCoE." The skill auto-triggers from its
description and routes to the right depth (a quick description vs. a full assessment).
The same workflow is available to Copilot as a prompt file at
.github/prompts/cloud-operating-model.prompt.md.
In VS Code, Visual Studio, or a JetBrains IDE with Copilot Chat, invoke it on demand with
/cloud-operating-model. It reuses the same cloud-operating-model/references/* and assets/*
files, so the two editions stay in sync from one source of truth.
Vendor framework specifics evolve (Azure CAF renamed operating models in 2025; AWS guidance changes). When current, citable wording is needed, the skill verifies against live documentation sources (Microsoft Learn, AWS docs) where available, and attributes named frameworks (Azure CAF, AWS, Gartner, Team Topologies) accurately. The archetype catalog stands on its own without them.
A companion skill for the enterprise architecture function lives in its own repo: enterprise-architecture-team.
MIT © 2026 Chris Horne