Skip to main content
CloudEval turns cloud context and infrastructure-as-code into reports, diagrams, grounded answers, and automation-ready artifacts across the browser, terminal, CI, and MCP-compatible agents. Use it when you need 650+ evaluation signals, shareable reports, a GitHub review workflow, and agent-ready context without rebuilding the review stack yourself.
Want to inspect the workflow before connecting anything private? Start with the public sample, then move into CLI, Web, GitHub, or MCP setup.

​
What you get

CLI and Web workspace

Run reports, ask grounded questions, open deeplinks, and keep the browser for visual review.

650+ evaluation signals

Turn cloud and IaC inputs into architecture, security, reliability, cost, diagram, and graph checks.

Any agent, anywhere

Use MCP, llms.txt, headless exports, and CLI output with ChatGPT, Claude, Codex, Cursor, and VS Code.

GitHub App and CI review

Sync repositories, run pull-request reviews, post comments, upload artifacts, and fail gates in CI.
CloudEval workspace with AI chat prompts next to an architecture diagram with issue badges, cost, report state, and share controls

​
Start with one path

Try public sample

Inspect a real example repo and demo PRs before setup.

Use the CLI

Install the CLI, run reports, ask questions, and automate reviews.

Review pull requests

Put CloudEval review comments, artifacts, and gates inside GitHub Actions.

Connect a source

Choose Cloud sync, a template, a workspace, or a GitHub-backed project.

Set up MCP

Give compatible agents CloudEval tools, context, and deeplinks.

Check coverage

See what is current, in progress, planned, and intentionally limited.

​
Next step

Open How CloudEval works to learn the platform model, or jump straight to the quickstart if you want hands-on setup.
Last modified on June 24, 2026