Skip to main content
By Prateek SinghPublished July 8, 2026
CloudEval light themed Quick Project Creation dialog importing files from a GitHub-hosted ARM template URL after github.com is changed to cloudeval.ai

The hook

GitHub URL to cloud diagram generation in less than a minute: keep the repository path, swap github.com for cloudeval.ai, and open a CloudEval project with diagrams and reports ready to review.

Many infrastructure reviews begin with a repository URL. Someone sends an Azure Quickstart template, an ARM deployment file, or a sample repository path, and the first reviewer has to decide whether the infrastructure is worth using, changing, or sending into production. The slow path is familiar: download the JSON, open another tool, create a project, upload the file, wait for parsing, then start preparing diagrams and notes for the actual review. CloudEval removes that context switch. For supported GitHub-hosted ARM template URLs, the repository path can become the project creation path. The shortcut is simple: keep everything after the host exactly the same, and replace only the domain. 
https://github.com/Azure/azure-quickstart-templates/blob/master/quickstarts/microsoft.compute/1vm-2nics-2subnets-1vnet/azuredeploy.json
Press Enter after the host swap. CloudEval opens the GitHub-style path as a quick project source, lets you review the import details, and then prepares the workspace for diagrams, reports, and exportable review documentation.

Import the source

Use a GitHub-hosted ARM JSON template as the project seed.

Inspect the diagram

Turn resource declarations and dependencies into an interactive architecture view.

Run reports

Generate cost and architecture reports from the same project workspace.

​
What CloudEval reads from GitHub

The URL shortcut is intentionally simple, but CloudEval treats the URL as an infrastructure source rather than a web page. For the single-template flow, the source should be an ARM JSON template that CloudEval can read from GitHub. Common GitHub URL shapes are supported, including blob, tree, and raw GitHub content URLs when they point to supported template content.
CloudEval Quick Project Creation progress state showing a consumed GitHub ARM template URL and import stages for diagram and report preparation
The strongest supported path today is Azure ARM JSON.
  • Use a public Azure Quickstart ARM template when you want to inspect a known sample.
  • Use a GitHub-hosted ARM JSON file when your own repository is already public or accessible.
  • Use a raw GitHub content URL when you want to point directly to the template bytes.
  • Add a parameters URL when the template needs one and auto-detection is not enough.
If your source of truth is Bicep, compile it to ARM JSON first, then import the generated ARM JSON file. For linked templates, nested templates, or private repositories, use the GitHub repository sync workflow instead of a one-off URL.
Use it for the fastest browser path from a GitHub-hosted ARM JSON template to a CloudEval project. It works best when the source is a single template and reviewers need a quick diagram/report workspace.
Use repository sync for private repositories, multi-file IaC projects, linked templates, branch provenance, commit tracking, and push-triggered refresh through the CloudEval GitHub App.
Use upload when the template is generated locally, not reachable by URL, or contains source details you do not want to expose through a public GitHub URL.

​
From URL to project workspace

After CloudEval resolves the GitHub source, it creates a project workspace around the infrastructure instead of treating the JSON as a throwaway upload.
1

Resolve the GitHub-style path

CloudEval normalizes the URL and reads the supported ARM JSON content from GitHub.
2

Create the project shell

The project receives a name, provider context, source reference, and source snapshot.
3

Detect parameters

If the GitHub URL follows a recognizable pattern, CloudEval attempts to find a matching parameters file from the same repository.
4

Parse resources and relationships

The template is interpreted as infrastructure code so resources, dependencies, and topology signals can be reviewed visually.
5

Open the workspace

CloudEval redirects into a project workspace where source files, diagram state, report status, and review surfaces live together.
CloudEval project workspace after GitHub URL import showing project files, report cards, code editor, validation count, cost, and Well-Architected status

​
The diagram is the first review surface

ARM templates can be hard to review directly because the architecture is spread across resources, parameters, variables, nested deployments, and implicit relationships. The generated diagram gives reviewers a shared visual model before they get lost in raw JSON.
CloudEval architecture diagram showing Azure resources and relationships in an interactive workspace
The diagram helps answer practical review questions:
  • What resources are in this template?
  • Which services depend on each other?
  • Does the network shape match the architecture we expected?
  • Are public-facing resources obvious before deployment?
  • Does the template need to be split into smaller reviewable pieces?

​
Reports turn the project into documentation

The project becomes useful to teams when the diagram and source context are paired with generated reports. CloudEval reports attach review output to the same project. Depending on the available evidence, reports can include architecture posture, cost context, resource inventory, security and compliance signals, assumptions, warnings, freshness, and suggested next actions.
CloudEval report Brief tab with posture, critical issues, run rate, savings, maturity, and freshness metrics

Brief

Executive readout for posture, urgent risks, run rate, maturity, freshness, and confidence.

Action Plan

Prioritized findings with effort, owner, impact, confidence, and suggested action.

Architecture

Dependency context, affected resources, and Well-Architected review signals.

Evidence

Source artifacts, assumptions, warnings, freshness, history, and export context.
CloudEval report overview showing posture score, critical issues, monthly run rate, savings, maturity, priority decisions, and evidence freshness
That is why this flow is more than “generate a diagram from GitHub.” It is URL-to-project, project-to-diagram, and project-to-report in one workspace.

​
What “prepares documentation” means

CloudEval does not just render a static picture. It prepares the review material teams usually need for architecture documentation and handoff. From the imported GitHub source, the project can produce:
  • An interactive architecture diagram for exploration.
  • A source-aware workspace with the imported template and related context.
  • Report tabs for engineering, platform, security, and stakeholder review.
  • Resource-level inventory tied back to findings and affected infrastructure.
  • Exportable report packets for meetings, audit handoff, or internal review.
  • Markdown, PDF, or JSON report exports where available.
  • Agent-ready context for Codex, Cursor, Claude, VS Code, and MCP workflows.
CloudEval issues list with filters, issue summaries, selected finding details, effort, confidence, and suggested actions

​
Shortcut versus durable sync

The domain-change shortcut is optimized for speed. Repository sync is optimized for repeatable source-of-truth workflows.
1

Open a GitHub ARM template URL

Start from the public or accessible GitHub URL your team already shared.
2

Replace only the host

Keep the repository owner, repo, branch, and file path exactly as they are. Swap github.com for cloudeval.ai, then press Enter.
3

Review and create

Confirm project name, provider, template URL, and optional parameters URL.
4

Inspect the workspace

Open the generated diagram, run reports, and export the review packet.
No for the initial GitHub URL import. The shortcut works from infrastructure code. Connect Azure separately when you want live environment sync.
Use GitHub repository sync for private source-of-truth workflows. The quick URL path is best for public or accessible URLs CloudEval can read.
No. The URL creates the project seed. Reports are generated from the project workspace after source import and setup are ready.

​
The takeaway

The shortest CloudEval workflow can start with one small habit: 
https://github.com/{owner}/{repo}/... -> https://cloudeval.ai/{owner}/{repo}/...
That change turns a GitHub-hosted infrastructure template into a CloudEval project creation path. From there, CloudEval consumes the ARM JSON, prepares the workspace, generates the diagram, runs reports, and gives the team exportable documentation for review.

Try the GitHub URL flow

Create a project directly from a GitHub-hosted ARM template URL.

Run reports

Generate cost and architecture reports from the project workspace.

Sync a repository

Use the CloudEval GitHub App when the repository should remain the source of truth.
Last modified on July 8, 2026