# cloud-audit

A read-only AWS scanner wired into your coding agent: attack-chain rules, IAM escalation routes and a blast-radius map per finding.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 73
- Author: gebalamariusz
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/gebalamariusz/cloud-audit
- Updated: 2026-10-14
- HTML: https://affpapa.org/en/nevibecode/mcp/cloud-audit

## What it does

This MCP server gives the agent a scanner for AWS accounts. It carries 110 checks, 64 known ways to climb the IAM privilege ladder, and 31 rules for spotting attack chains. Nothing is written to the account you point it at. Setup goes through `uvx cloud-audit` and the same entry works in Claude Code, Codex, Cursor, GitHub Copilot, Windsurf, Cline and Zed.

\`agent-blast` handles Bedrock Agents and AgentCore. For each agent it maps the roles sitting behind the tools, knowledge bases and gateways, then splits the result by threat model: the attacker holds a role's credentials, or they only got a prompt injected through a document or a ticket. Escalation paths are backed by the IAM policy simulator. Each finding arrives with an AWS CLI command and Terraform code that close it.

No account is needed to try it: `cloud-audit demo --save demo.json` builds a report from an invented account using the same engines. A real scan wants AWS credentials, and the AWS-managed `SecurityAudit` policy is enough for every check. The simulator reads identity policies, permissions boundaries and SCPs. Resource-based policies on the targets, RCPs and VPC endpoint policies fall outside it, so those edges of the report read friendlier than reality.

## Requirements
- Needs: Python and uv (commands run via uvx)

## Install

### Claude Code (native platform)

**Platform CLI**
```bash
claude mcp add cloud-audit -- uvx cloud-audit
```
uvx runs the command named after the package; if the README names another one, use uvx --from cloud-audit <command>

### Codex

**Platform CLI**
```bash
codex mcp add cloud-audit -- uvx cloud-audit
```
uvx runs the command named after the package; if the README names another one, use uvx --from cloud-audit <command>

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "cloud-audit": {
            "command": "uvx",
            "args": [
                "cloud-audit"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from cloud-audit <command>

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "cloud-audit": {
            "type": "stdio",
            "command": "uvx",
            "args": [
                "cloud-audit"
            ]
        }
    }
}
```
uvx runs the command named after the package; if the README names another one, use uvx --from cloud-audit <command>

### Gemini CLI

**Platform CLI**
```bash
gemini mcp add cloud-audit uvx cloud-audit
```
uvx runs the command named after the package; if the README names another one, use uvx --from cloud-audit <command>

## When to use it
- trace attack chains that end in account takeover
- see what a prompt-injected Bedrock agent can reach
- measure the blast radius of one role from a saved report
- grab the AWS CLI and Terraform fix for each finding
- audit a sample account before touching production credentials

## Example prompts
- Scan my AWS account for IAM escalation paths and walk me through the attack chains
- What can the hijacked `support-bot` agent reach? Run `agent-blast` with simulator proof
- Work out the blast radius of `arn:aws:iam::123456789012:role/support-bot-ticket-role` from `demo.json`

## Similar
- [REA](https://affpapa.org/en/nevibecode/mcp/rea.md) — MCP server · ★ 386989
- [World Monitor](https://affpapa.org/en/nevibecode/mcp/world-monitor.md) — MCP server · ★ 89408
- [mcp](https://affpapa.org/en/nevibecode/mcp/kubeshark-mcp.md) — MCP server · ★ 12109
- [Robin: AI-Powered Dark Web OSINT](https://affpapa.org/en/nevibecode/mcp/robin-ai-powered-dark-web-osint.md) — MCP server · ★ 7658
- [jshookmcp](https://affpapa.org/en/nevibecode/mcp/jshookmcp.md) — MCP server · ★ 2061
- [OpenOSINT](https://affpapa.org/en/nevibecode/mcp/openosint.md) — MCP server · ★ 1802

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md