gemini-skill

概述

Gemini-Skill 是一个通过 Chrome DevTools Protocol (CDP) 控制 Google Gemini 网页界面的综合 MCP 服务器。它提供对 Gemini 功能的自动化访问，包括 AI 图像生成及高清下载、多轮对话、图像上传生成以及从会话中提取图像。服务器采用守护进程模式，在后台维护浏览器实例，实现快速后续请求而无需每次重新启动浏览器。

试试问 AI

装完之后，这里有 4 个你可以让 AI 做的事：

• gemini_generate_imagegemini_generate_image(prompt, newSession, referenceImages, fullSize, timeout)

  Generate an image through Gemini AI with prompt

  试试:Generate an image of a sunset over the ocean试试:Create a cartoon drawing of a robot with a hat
• gemini_new_chatgemini_new_chat()

  Start a new blank conversation with Gemini

  试试:Start a new chat with Gemini试试:Reset the conversation to start fresh
• gemini_temp_chatgemini_temp_chat()

  Enter temporary conversation mode with Gemini

  试试:Create a temporary chat session试试:Start an ephemeral conversation
• gemini_switch_modelgemini_switch_model(model)

  Switch between different Gemini models

  试试:Switch to the pro model for better quality试试:Change to the quick model for faster responses
• gemini_send_messagegemini_send_message(message, timeout)

  Send a text message to Gemini and wait for a reply

  试试:Ask Gemini about the history of artificial intelligence试试:Explain quantum computing to me like I'm 5
• gemini_upload_imagesgemini_upload_images(images)

  Upload images to Gemini as reference for image generation

  试试:Upload these images as reference for generating similar images试试:Use this photo as a reference for a new design
• gemini_get_imagesgemini_get_images()

  Retrieve all image metadata from the current conversation

  试试:Show me all images in this conversation试试:List all images that have been shared
• gemini_extract_imagegemini_extract_image(imageUrl)

  Extract an image as base64 and save it locally

  试试:Extract this image as base64 and save it试试:Download the image from this URL
• gemini_download_full_size_imagegemini_download_full_size_image(index)

  Download the full-size high-resolution version of an image

  试试:Download the full-size version of the first image试试:Get the highest resolution of the last shared image
• gemini_share_latest_imagegemini_share_latest_image(index, timeout, copyToClipboard, closeDialog)

  Create a public share link for the latest image

  试试:Share the latest image and get a public link试试:Create a shareable link for the second image in the chat
• gemini_get_all_text_responsesgemini_get_all_text_responses()

  Get all text responses from the current conversation

  试试:Show me all the text responses in this chat试试:List everything Gemini has said in our conversation
• gemini_get_latest_text_responsegemini_get_latest_text_response()

  Get the latest text response from Gemini

  试试:What was the last response from Gemini?试试:Show me the most recent message from the AI

可对比工具

安装

git clone https://github.com/WJZ-P/gemini-skill.git
cd gemini-skill
npm install

BROWSER_DEBUG_PORT=40821
BROWSER_HEADLESS=false
DAEMON_TTL_MS=1800000
OUTPUT_DIR=./gemini-image

{
  "mcpServers": {
    "gemini": {
      "command": "node",
      "args": ["<gemini-skill绝对路径>/src/mcp-server.js"]
    }
  }
}

gemini-skill 对比