execute_browser_task

Automate browser actions by executing tasks described in plain text, enabling web automation through the Browser-use MCP Server.

Instructions

执行浏览器自动化任务

Input Schema

TableJSON Schema

Name	Required	Description	Default
`task`	Yes	要执行的任务描述

Implementation Reference

src/index.ts:94-108 (handler)

The core handler function that executes the browser task by running a shell command to activate a Python virtual environment and execute 'app.py --task "${task}"'. This is the exact implementation of the tool logic.

  private async executeBrowserTask(task: string): Promise<string> {
    const command = `cd /Users/charlesqin/PycharmProjects/PythonProject && \
source /Users/charlesqin/PycharmProjects/PythonProject/.venv/bin/activate && \
python app.py --task "${task}"`;

    try {
      const { stdout, stderr } = await execAsync(command);
      if (stderr) {
        console.error('执行命令时出现警告:', stderr);
      }
      return stdout || '任务执行完成，但没有输出结果';
    } catch (error: any) {
      throw new Error(`执行命令失败: ${error?.message || '未知错误'}`);
    }
  }

src/index.ts:60-91 (registration)

Registers the CallToolRequestSchema handler, which validates the tool name 'execute_browser_task', extracts the 'task' argument, calls the core handler, and formats the response or error.

this.server.setRequestHandler(CallToolRequestSchema, async (request) => {
  if (request.params.name !== 'execute_browser_task') {
    throw new McpError(
      ErrorCode.MethodNotFound,
      `未知工具: ${request.params.name}`
    );
  }

  const { task } = request.params.arguments as { task: string };

  try {
    const result = await this.executeBrowserTask(task);
    return {
      content: [
        {
          type: 'text',
          text: result,
        },
      ],
    };
  } catch (error: any) {
    return {
      content: [
        {
          type: 'text',
          text: `执行出错: ${error?.message || '未知错误'}`,
        },
      ],
      isError: true,
    };
  }
});

src/index.ts:41-58 (registration)

Registers the tool 'execute_browser_task' in the ListToolsRequestSchema response, including its description and input schema requiring a 'task' string.

this.server.setRequestHandler(ListToolsRequestSchema, async () => ({
  tools: [
    {
      name: 'execute_browser_task',
      description: '执行浏览器自动化任务',
      inputSchema: {
        type: 'object',
        properties: {
          task: {
            type: 'string',
            description: '要执行的任务描述',
          },
        },
        required: ['task'],
      },
    },
  ],
}));

src/index.ts:46-55 (schema)

Defines the input schema for the tool: an object with a required 'task' property of type string.

inputSchema: {
  type: 'object',
  properties: {
    task: {
      type: 'string',
      description: '要执行的任务描述',
    },
  },
  required: ['task'],
},

Browser-use MCP Server