Can the CLI export diagrams?

Yes. The Cloudeval CLI includes project diagram export workflows for architecture and dependency layouts.

Can the CLI ask architecture questions?

Yes. `cloudeval ask` supports project questions against available Cloudeval project context.

Can I use the CLI in automation?

Yes. Use non-interactive flags, doctor checks, and scoped access keys for automation-friendly workflows.

Does a green GitHub job prove a review ran?

No. Confirm that the evaluation step executed and inspect its findings, gate result and artifacts.

Can the action create architecture diagrams?

It can export diagrams and reports for supported project inputs. PR diagram diffs are not a current claim.

Does the workflow generate a diagram diff for every PR?

It can generate a fresh diagram or report artifact for the reviewed source. A PR diagram-diff workflow is not part of the current documented capability.

Answers

How do I generate architecture diagrams in GitHub Actions?

Quick answer

Use the Cloudeval GitHub Action to create a project review from supported infrastructure source, export diagrams and reports, and post the result to a pull request.

Configure scoped access and gates; a green workflow alone is not proof that an evaluation ran.

Install the Cloudeval CLI Create a Cloudeval project

Start here

Use this to

Move architecture review into the pull request with repeatable inputs, findings and exports.

Best source to use

Check the CLI overview and command reference before scripting automation, then run one project command by hand.

What you need

Cloudeval CLI installed.
Cloudeval login or scoped access key.
A project source or project ID.

What you get

Project creation or inspection from the terminal.
Reports, questions, sessions, or exports from CLI commands.
MCP serving when agents need tool access.

Get started

1
Install the Cloudeval CLI from the public install flow.
2
Run setup, login, status, and doctor commands before relying on automation.
3
Create or inspect Cloudeval Azure projects from the terminal.
4
Run reports or ask project questions with project IDs.
5
Export architecture or dependency diagrams when a headless workflow needs image files.
6
Configure MCP serving when an agent needs tool access through the CLI.
7
Define the source root and entry template in .cloudeval/config.yaml.
8
Run the action with a scoped access key and project ID.
9
Review the PR comment, optional Check Run or SARIF output, and artifacts.
10
Keep skipped checks and unsupported areas visible as not assessed.
11
Keep the source revision, gate result and exported artifact together; PR diagram diffs are not part of this workflow.

Helpful docs

CLI overview CLI command reference MCP setup support docs

When this works best

The GitHub workflow reviews supported Azure inputs and AWS CloudFormation in static beta.

PR diagram diffs and native repository scanning are not claimed as current capabilities.

AI Guide

Turn a diagram into a guided investigation

When a written answer is not enough, AI Guide walks through the architecture or dependency view, highlights the resources under discussion, and keeps the explanation, evidence, and next check together.

Personas adjust the emphasis for executive, architecture, security, or operations reviews. They do not change the available project evidence.

See how AI Guide works

Cloudeval Playground Architecture view with AI Guide entry point and a generated resource diagram — Real Playground capture. A guide explains available project evidence; it does not prove runtime traffic, health, or ownership.

Decision guide

Check	Cloudeval path	Note
Setup	Install, login, run doctor/status.	Do this before automation.
Project work	Create, open, inspect, or link a project.	Use project IDs for repeatable workflows.
Reports and AI	Run reports and ask project questions.	Prefer specific prompts.
Exports and agents	Export diagrams or serve MCP.	Use scoped credentials in CI.

Put the architecture review in the pull request

Keep the source root and entry template in .cloudeval/config.yaml, run the Cloudeval Action with a scoped access key and project ID, then inspect the PR comment, gate result and downloaded artifacts.

Terminal and agent workflow

This command creates Cloudeval projects from supported sources.
shell
```
cloudeval projects create
```
This command runs report workflows for a project.
shell
```
cloudeval reports run
```
This command sends one-shot project questions.
shell
```
cloudeval ask
```
This command exports architecture or dependency diagram images.
shell
```
cloudeval projects export-diagram
```
This command supports stdio MCP workflows where configured.
shell
```
cloudeval mcp serve
```

How to trust the result

Run one command manually before putting it into CI or an agent workflow.
Keep project IDs, account details, IPs, and resource names out of public exports unless they are sanitized.
Link exported diagrams back to the Cloudeval project or report when reviewers need details.

Frequently asked questions

How do I generate architecture diagrams in GitHub Actions?How does an MCP server help with cloud architecture review?How do I convert an ARM template to an architecture diagram?

Start here

Use this to

Move architecture review into the pull request with repeatable inputs, findings and exports.

Best source to use

Check the CLI overview and command reference before scripting automation, then run one project command by hand.

What you need

Cloudeval CLI installed.
Cloudeval login or scoped access key.
A project source or project ID.

What you get

Project creation or inspection from the terminal.
Reports, questions, sessions, or exports from CLI commands.
MCP serving when agents need tool access.

Get started

Install the Cloudeval CLI from the public install flow.

Run setup, login, status, and doctor commands before relying on automation.

Create or inspect Cloudeval Azure projects from the terminal.

Run reports or ask project questions with project IDs.

Export architecture or dependency diagrams when a headless workflow needs image files.

Configure MCP serving when an agent needs tool access through the CLI.

Define the source root and entry template in .cloudeval/config.yaml.

Run the action with a scoped access key and project ID.

Review the PR comment, optional Check Run or SARIF output, and artifacts.

Keep skipped checks and unsupported areas visible as not assessed.

Keep the source revision, gate result and exported artifact together; PR diagram diffs are not part of this workflow.

Turn a diagram into a guided investigation

Personas adjust the emphasis for executive, architecture, security, or operations reviews. They do not change the available project evidence.

See how AI Guide works

Decision guide

Check	Cloudeval path	Note
Setup	Install, login, run doctor/status.	Do this before automation.
Project work	Create, open, inspect, or link a project.	Use project IDs for repeatable workflows.
Reports and AI	Run reports and ask project questions.	Prefer specific prompts.
Exports and agents	Export diagrams or serve MCP.	Use scoped credentials in CI.

Terminal and agent workflow

This command creates Cloudeval projects from supported sources.

shell

cloudeval projects create

This command runs report workflows for a project.

shell

cloudeval reports run

This command sends one-shot project questions.

shell

cloudeval ask

This command exports architecture or dependency diagram images.

shell

cloudeval projects export-diagram

This command supports stdio MCP workflows where configured.

shell

cloudeval mcp serve

How do I generate architecture diagrams in GitHub Actions?

Start here

What you need

What you get

Get started

Helpful docs

When this works best

Turn a diagram into a guided investigation

Decision guide

Put the architecture review in the pull request

Terminal and agent workflow

How to trust the result

Frequently asked questions

Related pages

How do I generate architecture diagrams in GitHub Actions?

Start here

What you need

What you get

Get started

Helpful docs

When this works best

Turn a diagram into a guided investigation

Decision guide

Put the architecture review in the pull request

Terminal and agent workflow

How to trust the result

Frequently asked questions

Related pages