Browser Pool MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| BROWSER_POOL_DEBUG | No | Set 1 for debug logging to debug.log | 0 |
| BROWSER_POOL_BASE_PORT | No | Starting port for browser instances | 9000 |
| BROWSER_POOL_MAX_INSTANCES | No | Max concurrent browsers (≈300MB RAM each) | 10 |
| BROWSER_POOL_INSTANCE_TIMEOUT_MS | No | Idle timeout (30 min) | 1800000 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| browser_navigateC | 导航到 URL |
| browser_navigate_backD | 后退 |
| browser_navigate_forwardD | 前进 |
| browser_snapshotB | 获取页面可访问性快照（结构化，适合 LLM 阅读） |
| browser_take_screenshotC | 截图（默认 fullPage=false） |
| browser_screenshotC | 截图（别名） |
| browser_console_messagesC | 获取控制台消息 |
| browser_network_requestsC | 获取网络请求活动 |
| browser_get_urlC | 获取当前页面 URL |
| browser_clickC | 点击元素（element = a11y 引用，ref = CSS 选择器） |
| browser_typeC | 向元素输入文本 |
| browser_press_keyC | 按下键盘按键 |
| browser_hoverC | 悬停在元素上 |
| browser_select_optionC | 在下拉框中选择选项 |
| browser_dragC | 把元素 A 拖到元素 B |
| browser_dropB | 放下被拖拽的元素 |
| browser_evaluateC | 在页面中执行 JavaScript（函数字符串） |
| browser_wait_forD | 等待条件 |
| browser_tabsC | 管理标签页（action: list|new|close|select） |
| browser_resizeC | 调整浏览器窗口大小 |
| browser_handle_dialogC | 处理 alert/confirm 弹窗 |
| browser_file_uploadD | 上传文件 |
| browser_closeB | 关闭该 agent 的浏览器 |
| pool_statusA | 查看浏览器池状态（按 agent） |
| pool_testD | 测试异步响应 |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 25 tools
Most tools have distinct purposes, but browser_take_screenshot and browser_screenshot are explicit aliases, causing ambiguity. Additionally, browser_drag and browser_drop could be confused as both relate to drag-and-drop operations.
The majority of tools follow a browser_verb pattern (e.g., browser_navigate, browser_click), but there are exceptions like browser_snapshot, browser_console_messages, and browser_network_requests which use nouns. The pair browser_take_screenshot and browser_screenshot also breaks consistency.
At 25 tools, the server is on the heavy side but still plausible for a comprehensive browser automation tool. However, the presence of a duplicate screenshot tool and a test tool (pool_test) suggests some redundancy.
The tool set covers a wide range of browser actions: navigation, interaction, screenshots, snapshots, tab management, dialog handling, file upload, and more. Minor gaps exist (e.g., no explicit cookie management or scrolling) but core workflows are well covered.