Skip to content

Guide · Agent skill

How to use the Multi-Cloud Architecture Advisor agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

The multi-cloud-architecture skill gives an AI assistant a reference framework for designing applications that span Amazon AWS, Microsoft Azure, Google GCP and Oracle OCI. It is written for architects and engineers who are still planning a system and want a quick side-by-side of equivalent services. After this guide you will know what the skill does, how to install it, and how to ask for help with natural requests.

Multi-Cloud Architecture Advisorwshobson/agents · MIT

Quick start

  1. 1Download the ZIP. Download the multi-cloud-architecture ZIP file from this page.
  2. 2Unzip the folder. Extract the archive so the skill folder contains SKILL.md and the references folder.
  3. 3Place it in a skills folder. Move the folder into ~/.claude/skills/ for personal use, or into .claude/skills/ inside your project.
  4. 4Or upload it in the Claude apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the skill.
  5. 5Ask a natural question. Describe your system and the providers you are considering, such as asking for DR options between AWS and Azure.
  6. 6Review and verify. Check the suggested services and any cost figures against current provider documentation before deciding.

What is an agent skill and what does this one do?

An agent skill is a folder of instructions and reference files that an AI assistant loads when your request matches the skill's purpose. This one is defined in a SKILL.md file and supported by two reference files, references/multi-cloud-patterns.md and references/service-comparison.md.

Once it is active, the agent can compare equivalent services across AWS, Azure, GCP and OCI, for example EC2, Virtual Machines, Compute Engine and Compute for general-purpose VMs. It can also suggest architecture patterns, outline a four-phase migration plan (assessment, pilot, migration, optimization) and list cost-saving ideas.

  • Compare compute, storage and database services across the four providers
  • Recommend patterns such as Primary/DR pairing, best-of-breed service mix, or geographic distribution
  • Suggest cloud-agnostic building blocks such as Kubernetes, PostgreSQL and Terraform/OpenTofu
  • Draft a phased migration plan and cost optimization checklist

How do you install the multi-cloud-architecture skill?

Download the ZIP file from this page and unzip it. You can then place the folder in one of two locations: your personal skills folder at ~/.claude/skills/, or a project-specific .claude/skills/ folder inside your project. A personal install is available in every project; a project install applies only to that project.

If you use the Claude apps instead of a file system, you can upload the skill under Settings → Capabilities → Skills.

  • Personal install: unzip into ~/.claude/skills/
  • Project install: unzip into .claude/skills/ in your project folder
  • Claude apps: Settings → Capabilities → Skills, then upload
~/.claude/skills/multi-cloud-architecture/

How do you trigger the skill with a natural request?

You do not need a special command. The skill's description tells the agent to use it when you are building multi-cloud systems, avoiding vendor lock-in, or leveraging best-of-breed services from multiple providers. Plain requests that mention these goals should be enough.

Be specific about your workloads and the providers you are considering. The skill's comparison tables work best when the agent knows which service category you need.

  • Ask about a service: "What is the AWS, Azure, GCP and OCI equivalent of a managed Kubernetes cluster?"
  • Ask about a pattern: "Design a disaster recovery setup where the primary runs on AWS and the standby runs on Azure."
  • Ask about lock-in: "How can I keep this app portable across clouds without giving up managed databases?"
  • Ask for a plan: "Outline a phased migration from one cloud to two, with a pilot first."

What kinds of requests give useful answers?

The skill is strongest on early-stage design questions. Its patterns cover active-active regional splits, primary/DR pairing, best-of-breed mixes and portable platform baselines. Its cost section lists pricing models per provider, such as on-demand, reserved, spot and preemptible options.

Treat its output as a starting point for discussion. The files state general figures such as 30-70% savings for reserved or committed capacity without sources, so check them against your provider's current pricing.

  • Good fit: choosing services for a new system, comparing lock-in trade-offs, drafting DR options
  • Weaker fit: exact prices, current feature availability, or provider-specific configuration steps

What does the skill need, and does it run code or use the network?

The skill lists no required dependencies. It does not run scripts and does not use the network. It works from its included tables and written guidance rather than connecting to your cloud accounts.

Because it contains no provisioning scripts or provider-specific instructions, it cannot deploy anything for you. Teams that only use one cloud and need hands-on deployment help will not get much from it.

  • Requires: nothing listed
  • Runs scripts: no
  • Uses network: no
  • Does not access your AWS, Azure, GCP or OCI accounts

What should you watch out for?

Cloud offerings change, so verify service names, feature lists and prices with each provider before you commit to a design. Data sovereignty is mentioned only briefly in the skill; for regulated or safety-critical decisions, confirm requirements with specialists.

The migration phases and best practices are general guidance. They do not replace a detailed assessment of your own workloads and dependencies.

  • Verify service tables and cost figures against current provider documentation
  • Do not treat data sovereignty guidance as compliance advice
  • Use the output to plan, then validate with your own team

What people use it for

  • Comparing managed Kubernetes options (EKS, AKS, GKE, OKE) before choosing a platform
  • Drafting an active-active or primary/DR design across two providers
  • Choosing which provider to use for analytics, enterprise identity or regulated transaction workloads
  • Outlining a phased migration plan with a pilot workload
  • Listing cost-optimization ideas to review with your finance or cloud team

Frequently asked questions

How do I install the multi-cloud-architecture skill?

Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use or into .claude/skills/ inside a project. In the Claude apps, you can instead upload it under Settings → Capabilities → Skills.

Does the multi-cloud-architecture skill require anything else to run?

The skill lists no required dependencies. It runs no scripts and uses no network access, so it works from its included reference files.

Is the multi-cloud-architecture skill free?

The skill is published under the MIT licence. Check the repository's licence file for the exact terms before reuse.

Can the skill deploy my infrastructure to AWS, Azure, GCP or OCI?

No. It is advisory and contains no provisioning scripts or provider-specific instructions. It does not connect to your cloud accounts.

How does this skill compare with a single-cloud or Terraform-only approach?

It focuses on design decisions across four providers, including service equivalents and patterns. It does not provide deployment code, so Terraform modules or a single-cloud guide are better for hands-on implementation.

Are the service comparison tables and cost figures up to date?

Not guaranteed. Cloud services and pricing change, and the files give general figures such as 30-70% savings without sources, so verify them with each provider.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Use this skill if you are early in planning a system that may span more than one cloud, and you want a quick, structured comparison of AWS, Azure, GCP and OCI services. Pick something else if you run on a single cloud and need hands-on deployment help, or if you need verified pricing or compliance advice.

Download the skill