# Cloudeval AI MCP server

> Evaluates Azure and AWS CloudFormation templates for cost, architecture, and security risks using an AI-ready terminal interface.

- Repository: https://github.com/ganakailabs/cloudeval-cli
- Page: https://hostedmcp.org/servers/ganakailabs/cloudeval-cli
- Transport: stdio
- Stars: 1
- Latest version: 0.38.5

## Install

### Claude Code

```sh
claude mcp add --env 'CLOUDEVAL_ACCESS_KEY=<YOUR_CLOUDEVAL_ACCESS_KEY>' cloudeval-cli -- npx -y @ganakailabs/cloudeval-cli
```

### Cursor (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "cloudeval-cli": {
      "command": "npx",
      "args": [
        "-y",
        "@ganakailabs/cloudeval-cli"
      ],
      "env": {
        "CLOUDEVAL_ACCESS_KEY": "<YOUR_CLOUDEVAL_ACCESS_KEY>"
      }
    }
  }
}
```

### VS Code (`.vscode/mcp.json`)

```json
{
  "servers": {
    "cloudeval-cli": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@ganakailabs/cloudeval-cli"
      ],
      "env": {
        "CLOUDEVAL_ACCESS_KEY": "<YOUR_CLOUDEVAL_ACCESS_KEY>"
      }
    }
  }
}
```

### VS Code CLI

```sh
code --add-mcp '{"name":"cloudeval-cli","type":"stdio","command":"npx","args":["-y","@ganakailabs/cloudeval-cli"],"env":{"CLOUDEVAL_ACCESS_KEY":"<YOUR_CLOUDEVAL_ACCESS_KEY>"}}'
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "cloudeval-cli": {
      "command": "npx",
      "args": [
        "-y",
        "@ganakailabs/cloudeval-cli"
      ],
      "env": {
        "CLOUDEVAL_ACCESS_KEY": "<YOUR_CLOUDEVAL_ACCESS_KEY>"
      }
    }
  }
}
```

### Codex

```sh
codex mcp add cloudeval-cli --env 'CLOUDEVAL_ACCESS_KEY=<YOUR_CLOUDEVAL_ACCESS_KEY>' -- npx -y @ganakailabs/cloudeval-cli
```

### Codex config.toml (`~/.codex/config.toml`)

```toml
[mcp_servers.cloudeval-cli]
command = "npx"
args = ["-y", "@ganakailabs/cloudeval-cli"]

[mcp_servers.cloudeval-cli.env]
CLOUDEVAL_ACCESS_KEY = "<YOUR_CLOUDEVAL_ACCESS_KEY>"
```

### Gemini CLI

```sh
gemini mcp add -s user -e 'CLOUDEVAL_ACCESS_KEY=<YOUR_CLOUDEVAL_ACCESS_KEY>' cloudeval-cli npx -- -y @ganakailabs/cloudeval-cli
```

### Gemini settings.json (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "cloudeval-cli": {
      "command": "npx",
      "args": [
        "-y",
        "@ganakailabs/cloudeval-cli"
      ],
      "env": {
        "CLOUDEVAL_ACCESS_KEY": "<YOUR_CLOUDEVAL_ACCESS_KEY>"
      }
    }
  }
}
```

## Environment variables

| Name | Required | Secret | Description |
|---|---|---|---|
| `CLOUDEVAL_ACCESS_KEY` | yes | yes | Cloudeval access key with permissions for the selected project and report reads. Create it in the Cloudeval app under Developer > API & CLI access keys. |
| `CLOUDEVAL_TELEMETRY` | no | no |  |

## FAQ

### Can I use Cloudeval with AI agents like Claude or Cursor?

Yes, it can be used as an MCP server for Codex, Cursor, Claude, VS Code, and any stdio JSON-RPC client.

### Does Cloudeval send my private cloud data to its servers?

Telemetry is curated and does not send raw prompts, command output, local paths, project identifiers, or cloud resource names.