ChenyuHeee/dsh-browser-playwright

专为DeepSeek Harness设计的快照优先Playwright浏览器自动化工具:通过可访问性树进行交互并具备稳定引用,每会话浏览器上下文,共17个browser_*工具。

Project Overview项目介绍

This is a native browser automation plugin built exclusively for DeepSeek Harness, powered by Playwright. It enables DSH agents to interact with real web pages by generating accessibility snapshots with stable element references, so there is no need to guess CSS selectors or dump full DOM trees to complete actions. Each DSH user session gets its own dedicated browser session, and the plugin supports tab management, durable PNG screenshot attachments, structured web content extraction, and gated JavaScript evaluation. To install the plugin, you can run the command dsh plugin --profile <your-profile-name> add dsh-browser-playwright directly from the DSH CLI.

The plugin exposes 17 model-facing tools with a default name prefix of browser_, covering every common browser action from navigation to closing the browser session. After every action the plugin generates a new fresh snapshot, so all element references are pulled from the most recent page state to avoid stale interaction errors. Snapshot output automatically excludes invisible content, caps the number of nodes and length of text to keep token usage manageable, and adds an explicit flag if content is truncated. This plugin is ideal for developers who need agents to autonomously complete web interaction, scraping, and operation tasks.

The plugin automatically probes for installed Chromium-family browsers, so you do not need to download a new browser binary to get started. If you prefer to use a Playwright-managed Chromium binary, you can run npx playwright-core install chromium to set it up, or use a mirror for restricted networks like mainland China. There are several known limitations: only Chromium is supported, purely visual custom widgets may not get element references, and cross-origin iframes do not support deep traversal. The plugin is released under the permissive open-source MIT license.

这是面向 DeepSeek Harness 开发的原生浏览器自动化插件,基于 Playwright 实现,可为 DSH 代理提供真实浏览器驱动能力。它通过生成带有稳定元素引用的可访问性快照完成交互,不需要猜测 CSS 选择器,也不会输出完整的 DOM 快照占用多余 tokens。每个 DSH 会话对应一个独立浏览器会话,支持标签页管理、持久化截图附件、结构化网页提取和受管控的 JavaScript 执行。安装可直接通过 DSH 命令行完成:dsh plugin --profile <名称> add dsh-browser-playwright。

插件提供了 17 个面向模型的工具,默认前缀为 browser_,涵盖导航、快照、点击、填充、按键、滚动、历史进退、等待、标签页管理、截图、提取、求值、关闭等全流程操作。每次操作后都会生成新的快照,元素引用始终来自最新的结果,保证交互准确性。快照会自动排除不可见内容,限制节点数量和文本长度,控制 token 消耗,截断时会在结果中标记。适合需要代理自主完成网页交互、信息爬取、网页操作任务的开发者使用。

插件会自动探测已安装的 Chromium 内核浏览器,无需额外下载浏览器,若需要使用 Playwright 管理的 Chromium,可执行 npx playwright-core install chromium 命令安装。它仅支持 Chromium 系列浏览器,不支持 Firefox 或 WebKit。存在一些已知限制,比如纯视觉绘制的控件可能无法生成元素引用,跨域 iframe 暂不支持深度遍历,开启 JavaScript 求值需要手动修改配置。它采用 MIT 许可开源使用。

Pre-install check安装前体检Compatibility · Security兼容性 · 安全性 1 warning1 项注意
  • Only 8 stars - very few users, little community feedback星标只有 8,几乎没人在用,遇到问题缺少社区反馈
DSH walks through these 9 checksDSH 会逐条核对这 9 项

Compatibility兼容性

  • DSH, Node, OS and profile requirementsDSH 版本 / Node 版本 / 操作系统 / profile 是否满足要求
  • External dependencies and runtimes (Electron / Python / Docker, ...)外部依赖与运行时(Electron / Python / Docker 等)是否齐备
  • Conflicts with installed plugins: command names, skill / tool names, ports, duplicate MCP registration与已装插件是否冲突:命令名、skill / tool 重名、端口占用、重复 MCP 注册

Security安全性

  • Repo matches the facts registered here; archived or abandoned?仓库是否与页面登记一致,是否归档或长期停更
  • Safety of preinstall / install / postinstall and install.sh / setup.ps1preinstall / install / postinstall 与 install.sh、setup.ps1 是否安全
  • curl|bash, download-then-execute, obfuscation, unrelated domains → stop immediatelycurl|bash、下载即执行、混淆代码、无关域名 → 立刻停止
  • Typosquatting or unmaintained packages among the new dependencies新增依赖里有没有 typosquatting 或无人维护的包
  • Requested permissions vs. what the feature actually needs申请了哪些权限、是否超出功能所需(filesystem / network / shell / clipboard)
  • Any sudo / admin requirement, plus uninstall and rollback是否要求 sudo / 管理员权限,以及卸载与回滚方式

Anything uncertain must be marked unknown with a note on how to confirm it. This site's signal screen is a static snapshot, not a security audit.拿不准的必须标「未知」并说明要我怎么确认。本站的信号筛查是静态快照,不能替代安全审计。

Or use CLI install (for developers)或使用命令行安装(适合开发者)

CLI Install命令行安装

dsh plugin --profile web add dsh-browser-playwright

把 ChenyuHeee/dsh-browser-playwright 加入你的 DSH 配置(web profile)即可启用。

READMEREADME

dsh-browser-playwright

Playwright-powered browser capability for DeepSeek Harness: the agent drives a real browser through accessibility snapshots with stable element refs — no CSS-selector guessing, no full-DOM dumps. One browser session per harness session, tabs, screenshots as durable image attachments, structured extraction, and gated JavaScript evaluation.

Install

dsh plugin --profile <name> add dsh-browser-playwright

The bundle mounts three rows: the ctx.browser seam (service), the Playwright provider (playwright), and the model-facing tool family (tool).

Browser binary

The provider probes chromium → chrome → msedge → edge channels automatically; any installed Chromium-family browser works with zero download. To use Playwright-managed Chromium instead:

npx playwright-core install chromium
# Restricted networks (e.g. mainland China):
PLAYWRIGHT_DOWNLOAD_HOST=https://cdn.npmmirror.com/binaries/playwright npx playwright-core install chromium

Or pin a binary: launch.executablePath / launch.channel in the plugin config.

Tools

The default toolPrefix is browser_. Every action returns a fresh snapshot, so refs always come from the latest result.

Tool Purpose
browser_navigate Open a URL (http/https only) and return the snapshot
browser_snapshot Snapshot the current page: URL, title, visible elements with ref= ids
browser_click Click the element with a ref from the latest snapshot
browser_fill Replace an input's value; selects match by option label
browser_press Press a key (Enter, Tab, Escape, ArrowDown, …) on an element
browser_scroll Scroll the page or bring a ref into view
browser_back / browser_forward History navigation
browser_wait Wait a bounded duration for lazy content
browser_tabs / browser_open_tab / browser_switch_tab / browser_close_tab Tab management
browser_screenshot PNG capture stored as a durable image attachment the model can view
browser_extract Structured extraction: natural-language instruction → one parsed JSON value
browser_evaluate Run one JavaScript expression in the page (disabled by default)
browser_close Close the session's browser; the next call opens a fresh one

Showing the opening section of the README — the full document lives in the repository以上为 README 开头摘要,完整文档在仓库内 · View the full README on GitHub →在 GitHub 查看完整 README →

← 上一个 Prev dsh-session-header 下一个 Next dsh-plugin-studio →