Updated May 12, 2026
How do I convert an ARM template to an architecture diagram?
Quick answer
Create a Cloudeval Azure project from an ARM JSON template.
Cloudeval turns the resources into architecture and dependency diagrams, then lets you run reports or ask questions about risks, costs, and relationships.

What you can expect
A practical path from source input to diagram, report context, docs, exports, and the first AI question to ask.
Start here
Use this to
Review an ARM template visually before or after deployment.
Best source to use
Use ARM JSON as the input. If the source is Bicep, compile it to ARM JSON first.
What you need
- A valid ARM JSON template or supported GitHub template URL.
- A Cloudeval account.
- Enough project context to run the reports you care about.
What you get
- A project diagram from ARM resources.
- Architecture, security, cost, Well-Architected, or ARM/unit-test findings where available.
- A shareable review path for pull requests or design reviews.
Fastest path to a useful result
- 1Create a new Azure project.
- 2Add the ARM JSON template or supported GitHub template URL.
- 3Open the architecture diagram.
- 4Run one report and review the highest-priority finding.
When this works best
Use this for Azure ARM JSON templates.
For Bicep, compile first; for Terraform or CloudFormation, confirm current support before committing the workflow.
Decision guide
| Check | Cloudeval path | Note |
|---|---|---|
| Input | ARM JSON template or supported GitHub URL. | Use compiled output for Bicep. |
| Diagram | Architecture and dependency views. | Use both views during review. |
| Evidence | Reports and project graph context. | Missing context should be treated as unknown, not safe. |
| Best use | Pre-deployment review and design discussion. | Good for pull requests and platform reviews. |
Who this is for
- Infrastructure engineers
- Azure consultants
- Platform teams reviewing ARM templates
- AI agents preparing IaC reviews
How it works
- 1Open Cloudeval and create a new project.
- 2Choose Azure as the provider and provide an ARM JSON template file or supported GitHub template URL.
- 3Let Cloudeval parse resources and relationships into project graph context.
- 4Open the architecture diagram for human-readable grouping and layout.
- 5Open the dependency view when you need a relationship-focused view.
- 6Run reports or ask questions to explain risks, costs, and architecture tradeoffs.
ARM template review before deployment
Before deploying an ARM template, load it into Cloudeval and check whether network, identity, compute, and data resources connect as expected.
- Then ask which resources introduce security, reliability, or cost concerns.
- The output can support a design review or pull request discussion.
Terminal and agent workflow
Use this command for terminal-driven project creation.
shellcloudeval projects create --template-file ./azuredeploy.json --name "ARM review" --provider azureUse this command after project creation when reports are needed.
shellcloudeval reports run --project <project-id> --type all --wait- MCP-compatible agents can use Cloudeval context to answer ARM review questions from supported clients.
How to trust the result
- Compare the diagram with the template resources before using it in a design review.
- Ask which relationships are missing when the diagram looks simpler than the template.
- Run reports before sharing the diagram so reviewers see risks, not just topology.
FAQ
Is a deployed Azure account required?
No.
- Infrastructure-as-code projects can be created from ARM template files or supported GitHub URLs.
- Cloud sync is a separate workflow.
Can Cloudeval review ARM templates for security?
Yes, Cloudeval can run architecture, security, Well-Architected, and ARM/unit-test reports where the required project context and checks are available.
Can I export the resulting diagram?
Yes.
Cloudeval supports browser and CLI diagram export workflows for project diagrams.
Related pages
Build this workflow in Cloudeval
Start with a Cloudeval Azure project, then connect diagrams, reports, source links, CLI automation, and MCP-compatible agent workflows around the same source of truth.
