一个MCP 服务器允许您创建 PowerPoint 演示文稿 。 由SlideSpeak提供动力,现在可以使用SlideSpeak MCP创建演示文稿. 自动报告,演示另一个幻灯片甲板。 从今天开始!

□ 用克洛德桌面

要使用 Claude 桌面, 请在您的 Claude desktop config.json 中添加以下内容 :

远程 MCP

这是运行MCP最简单的方式. 这个方法需要您在您的系统中安装节点.js.

(这里免费下载节点.js)

{\fn华文楷体\fs16\1cHE0E0E0}多克

这将允许您在自己的计算机上运行 MCP 服务器 。 这种方法需要将多克安装在您的系统中 。

(这里可以免费下载 Docker 桌面)

□ 获得 API 密钥

访问此页面以获取幻灯片语音的 API 密钥 : https://slidespeak.co/slidespeak-api/

□ 开发幻灯片Speak MCP

以下信息与幻灯片Speak MCP的开发有关. 不需要这些步骤来使用多程序程序。

QQ 构建 Docker 图像

这是用于本地测试的, 如果您想要发布一个新的嵌入容器检查“ 制作新版本” 部分 见下文。

发展

· 安装uv

建立虚拟环境并激活它

阴道

slidespeak-mcp

An MCP Server that allows you to create PowerPoint presentations. Powered by SlideSpeak, you can now create presentations using the SlideSpeak MCP. Automate reports, presentations an other slide decks. Start today!

Usage with Claude Desktop

To use this with Claude Desktop, add the following to your claude_desktop_config.json:

Remote MCP

This is the easiest way to run the MCP. This approach requires you to have Node.js installed on your system.

(Download Node.js for free here)

{
  "mcpServers": {
    "slidespeak": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://mcp.slidespeak.co/mcp",
        "--header",
        "Authorization: Bearer YOUR-SLIDESPEAK-API-KEY-HERE"
      ],
      "timeout": 300000
    }
  }
}

Docker

This will allow you to run the MCP Server on your own computer. This approach requires Docker to be installed on your system.

(Download Docker Desktop for free here)

{
  "mcpServers": {
    "slidespeak": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-e",
        "SLIDESPEAK_API_KEY",
        "slidespeak/slidespeak-mcp:latest"
      ],
      "env": {
        "SLIDESPEAK_API_KEY": "YOUR-SLIDESPEAK-API-KEY-HERE"
      }
    }
  }
}

Getting an API key

Visit this page in order to get an API key for Slidespeak: https://slidespeak.co/slidespeak-api/

Development of SlideSpeak MCP

The following information is related to development of the SlideSpeak MCP. These steps are not needed to use the MCP.

Building the Docker Image

This is for local testing, if you want to publish a new docker container check out the “Making a new version” section below.

docker build . -t slidespeak/slidespeak-mcp:TAG-HERE

Development

Install uv

curl -LsSf https://astral.sh/uv/install.sh | sh

Create virtual environment and activate it

uv venv source .venv/bin/activate

Install dependencies

uv pip install -r requirements.txt

Using the server directly without Docker

Add the following to your claude_desktop_config.json:

{
  "mcpServers": {
    "slidespeak": {
      "command": "/path/to/.local/bin/uv",
      "args": [
        "--directory",
        "/path/to/slidespeak-mcp",
        "run",
        "slidespeak.py"
      ],
      "env": {
        "SLIDESPEAK_API_KEY": "API-KEY-HERE"
      }
    }
  }
}

Making a new release

Version naming should be in the format of MAJOR.MINOR.PATCH (e.g., 1.0.0).

The version needs to be updated in the following files:

• pyproject.toml -> version
• slidespeak.py -> USER_AGENT

Make a new release in GitHub and tag it with the version number. This will trigger a GitHub Action. The release will be automatically built and pushed to Docker Hub.