Windows Automation MCP Server

Instructions

Implementation Reference

• src/tools/browser.js:217-233 (handler)

  The `screenshot` method that executes the `browser_screenshot` tool logic using Puppeteer's `page.screenshot()` to capture the page screenshot and save it to a file.

  async screenshot(savePath, fullPage = false, sessionId = 'default') {
    try {
      const page = this.pages.get(sessionId);
      if (!page) {
        return { success: false, error: '浏览器未启动' };
      }
  
      const timestamp = new Date().toISOString().replace(/[:.]/g, '-');
      const defaultPath = `screenshot-browser-${timestamp}.png`;
      const path = savePath || defaultPath;
  
      await page.screenshot({ path, fullPage });
  
      return { success: true, path, fullPage, message: '截图已保存' };
    } catch (error) {
      return { success: false, error: error.message };
    }

• src/tools/browser.js:75-86 (schema)

  The tool definition including the input schema for `browser_screenshot` in `getToolDefinitions()`.

  {
    name: 'browser_screenshot',
    description: '截取网页截图',
    inputSchema: {
      type: 'object',
      properties: {
        path: { type: 'string', description: '保存路径（可选）' },
        fullPage: { type: 'boolean', description: '是否整页截图（可选）' },
        sessionId: { type: 'string', description: '会话 ID（可选）' },
      },
    },
  },

• src/tools/browser.js:135-136 (registration)

  The switch case in `executeTool` method that registers and dispatches the `browser_screenshot` tool call to its handler.

  case 'browser_screenshot':
    return await this.screenshot(args.path, args.fullPage, args.sessionId);