MCP.so
ログイン

AWS Resources MCP Server

@MCP-Mirror

AWS Resources MCP Server について

Mirror of

基本情報

カテゴリ

クラウドとインフラ

ライセンス

MIT license

ランタイム

python

トランスポート

stdio

公開者

MCP-Mirror

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "baryhuang_mcp-server-aws-resources-python": {
      "command": "npx",
      "args": [
        "-y",
        "@smithery/cli",
        "install",
        "mcp-server-aws-resources-python",
        "--client",
        "claude"
      ]
    }
  }
}

ツール

1

Execute a boto3 code snippet to query AWS resources

概要

What is AWS Resources MCP Server?

A Model Context Protocol server that runs generated Python code to query any AWS resources through boto3. It is intended for users who want to interact with AWS services via natural language through an MCP client like Claude Desktop.

How to use AWS Resources MCP Server?

Install via Smithery (npx -y @smithery/cli install ...) or pull the Docker image from Docker Hub. Configure AWS credentials using environment variables (AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, AWS_DEFAULT_REGION, optional AWS_SESSION_TOKEN) or by mounting a local ~/.aws credentials file and setting AWS_PROFILE. Invoke the query_aws_resources tool with a code_snippet string containing boto3 code that sets a result variable.

Key features of AWS Resources MCP Server

  • Runs from a Docker image – no git clone required
  • Uses Python and boto3 for all AWS resource queries
  • Supports Linux/amd64, Linux/arm64, and Linux/arm/v7 platforms
  • Sandboxes code execution with a restricted set of allowed imports and built-ins
  • Accepts both access key and profile-based AWS authentication

Use cases of AWS Resources MCP Server

  • Query AWS resources (e.g., S3 buckets, CodePipeline deployments) through natural language
  • Debug and fix AWS configuration issues, such as DynamoDB permission errors
  • Retrieve operational data from AWS using generated boto3 code

FAQ from AWS Resources MCP Server

What AWS credentials are required?

AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, optionally AWS_SESSION_TOKEN, and AWS_DEFAULT_REGION. Alternatively, you can use a profile stored in ~/.aws/credentials by setting the AWS_PROFILE environment variable.

Is the server limited to read-only operations?

No. The author did not restrict operations to ReadOnly, so your IAM user role dictates what operations you can perform.

How do I run the server?

You can either pull and run the Docker image (docker pull buryhuang/mcp-server-aws-resources:latest) or clone the repository and run it locally with uv. For Claude Desktop, configure the MCP server with a Docker command and pass in the required environment variables.

What imports are allowed in the code snippet?

Only these imports are permitted: boto3, operator, json, datetime, and pytz. A set of built-in functions (e.g., len, max, sorted, dict) is also available.

Where does the executed code run?

The code runs inside a Docker container, which provides sandboxing and keeps execution environment clean and isolated.

コメント

「クラウドとインフラ」の他のコンテンツ