Model Context Protocol moderate risk

AWS

This MCP server grants Claude access to AWS services via the AWS CLI, enabling automation and management of AWS resources based on configured IAM permissions.

Connections & Capabilities

Connects To

GitHubDockerAWS

Capabilities

readwriteexecadmin

Quickstart

Install

pip install -e

Config

{
  "mcpServers": {
    "aws": {
      "command": "uvx",
      "args": ["aws-mcp"]
    }
  }
}

Exposed MCP Tools (2)

safe

aws_cli_help

Retrieves documentation for any AWS CLI command.

Read-only operation that only retrieves help documentation.

high

aws_cli_pipeline

Executes AWS CLI commands with optional pipes (e.g., jq, grep).

Allows execution of arbitrary AWS CLI commands, potentially leading to resource modification or data exfiltration depending on IAM permissions.

Safety Assessment

The server's safety depends heavily on the configured IAM policy. When using a least-privilege IAM role and Docker containerization, the risk is reduced. However, misconfigured IAM policies can lead to significant security risks.

IAM policy acts as the primary security boundary, limiting Claude's actions.
Docker containerization provides additional isolation (optional).
Sandbox mode can further restrict access.
Least-privilege IAM role is recommended.

Unrestricted IAM policies could allow Claude to perform destructive actions.
Direct access to AWS CLI exposes a wide range of services.
Improperly configured credentials can lead to unauthorized access.
Potential for command injection if input validation is insufficient.