@tokens/dsh-browser-use
@tokens/dsh-browser-use
TokensHarness 浏览器操作插件:让模型直接驱动一个可见的真实浏览器(导航、点击、填表、截图、读取页面结构)。
本插件不自研浏览器自动化——它是薄薄一层产品化封装:
浏览器操作由 microsoft/playwright-mcp(Playwright 官方 MCP server,随本插件打包)提供,共 24 个工具;
接入桥由上游
@deepseek-ai/dsh-mcp-client提供,工具以mcp__browser__*注册给模型;本插件只做参数翻译(用户友好的 Config → CLI 参数)与跨平台 spawn(Windows 经 ComSpec 包装,不依赖 npx 与网络)。
默认行为
有头模式:模型操作时弹出真实浏览器窗口,全程可见;
复用系统 Chrome:不下载独立 Chromium,安装包零体积负担;
持久化 profile:保留登录态,模型可以在你已登录的站点上工作。
Related MCP server: Cloudflare Playwright MCP
配置
- id: tokens-browser-use
name: '@tokens/dsh-browser-use'
config:
browser: chrome # chrome | msedge | chromium | firefox | webkit
headless: false # true 则后台运行不弹窗
isolated: false # true 则每次全新环境(不保留登录态)
caps: '' # 追加能力,如 "vision,pdf"
executablePath: '' # 浏览器可执行文件绝对路径,留空自动发现所有默认值集中在 identity.js,用户可经 profile 补丁按需覆盖,改配置不需重启以外的操作(装/卸插件需重启)。
开发
npm install
npm run check # 语法 + 6 项单测This server cannot be deployed
Maintenance
Related MCP Connectors
E2LLM gives your AI eyes and hands in a real browser: structured perception (SiFR) plus action.
Cloud browsers for AI agents: get past bot detection, sign in with MFA, run web tasks at scale.
1- openhelmOAuthai.openhelm
Autonomous cloud agent tasks: real browser + your tools, structured evidence-backed results.
Stealth web automation for AI agents. Login, signup, navigate, screenshot.
Related MCP Servers
- FlicenseNot gradedqualityCmaintenanceEnables AI assistants to control a browser through a set of tools, allowing them to perform web automation tasks like navigation, typing, clicking, and taking screenshots.-
- FlicenseNot gradedqualityCmaintenanceEnables AI assistants to control a browser through a set of tools, allowing them to perform web automation tasks like navigation, typing, clicking, and taking screenshots.-
- AlicenseAqualityDmaintenanceEnables AI agents to control a browser with 30 tools for navigation, interaction, extraction, and tab management, supporting human-like browser automation.365 npmMIT
- FlicenseNot gradedqualityCmaintenanceEnables AI assistants to control a browser through a set of tools, allowing them to perform web automation tasks like navigation, typing, clicking, and taking screenshots.-