opencli抖音自动化发布内容:用 OpenCLI 在创作中心发文章、图文与视频

SkillHub 社区 Skill @user_401e3080/opencli-douyin(v1.0.0,business-ops,无需 API Key)将抖音创作中心发布流程写入 SKILL.md:用 OpenCLI 控制已登录 creator.douyin.com,支持文章、图文、视频;覆盖 React 填表、Semi 封面上传、短信验证与可选 douyin 适配器命令。文章说明 SkillHub 安装、Node 21+ 与 OpenCLI 环境、典型 browser/适配器用法及登录态与页面改版等注意点。

Read More
猎聘搜索:在猎头端用 Agent 自动搜人并批量打招呼

猎聘搜索是 SkillHub 社区技能 @user_09d1a78f/liepin-guide(v1.0.0,business-ops),面向猎聘猎头端 lpt.liepin.com。用户在浏览器登录猎头账号后,由 Agent 通过浏览器自动化按招聘要求搜索候选人、按活跃度与匹配度筛选,并批量发送招呼;无需 API Key。安装可让 AI 按 skillhub.cn/install/skillhub.md 安装该坐标,或使用 skillhub install 并指定 Agent 的 skills 目录。适用猎头日常 sourcing,需注意登录态、平台规则与社区维护节奏。

Read More
browser-useV1:用自然语言驱动 AI 浏览器自动化

browser-useV1(@user_fe8721be/jmy004)是 SkillHub 上的 ai-agent 类 Skill,将 browser-use 的 CLI 与 TypeScript API 整理为 Agent 可读指南。开发者用自然语言描述打开页面、提取数据、填表等任务,由 LLM 决策浏览器操作,区别于命令式 playwright-cli。文章说明 Skill 定位、核心能力、skillhub 安装到 Cursor 等目录的步骤、API Key 配置,以及典型命令与代码示例,并提示域名限制与安全注意事项。

Read More
头条文章阅读:用浏览器自动化抓取多平台正文

头条文章阅读是 SkillHub 社区 Skill(@user_09d1a78f/toutiao-article-reade,v1.0.0,data-analysis),通过浏览器自动化阅读今日头条、微信公众号、知乎等链接,提取标题、作者、发布时间与正文,支持内容清洗与 AI 总结,标签 requires_api_key=false。文章说明其定位、核心能力,以及按 skillhub.cn/install/skillhub.md 用 skillhub install 与 --dir 安装到 Cursor 等客户端 skills 目录的步骤,并提示自动化与合规使用注意。

Read More
DeepSeek Harness 插件介绍:dsh-browser

`dsh-browser` 插件旨在解决 DeepSeek Harness (DSH) 浏览器自动化中易被识别的风险。它不依赖 CDP 协议,通过 Native Messaging 与扩展通信,并使用持久 Profile 保留登录态,从而降低反爬检测。核心功能包括文本化页面交互、丰富的浏览器控制工具集以及基于令牌的配对机制。安装过程包含插件安装、连接组件配置、手动加载扩展及配对。该工具适合在有反爬机制或需保持登录态的复杂网页交互场景中使用。

Read More
openclaw free-gemini:免 Google API Key 接入 Gemini 的 OpenClaw Skill

openclaw free-gemini(@user_627aed5d/gemini001,v1.0.0)是 SkillHub 社区 ai-agent 类 Skill,面向 OpenClaw 框架,通过内置路由实现无 Google API Key 接入 Gemini 全系列模型。Gemini 负责推理与规划,OpenClaw 负责调度、Edge 浏览器自动化、网页交互、重连与持久运行。安装需按 skillhub.cn 文档使用 skillhub install 并指定 Agent 的 skills 目录。适用快速试浏览器自动化与无人值守任务;官方注明需要梯子,且未在页面提供详细启动示例,需以安装后 SKILL 与 OpenClaw 文档为准。

Read More
网页操作办公助手(Browser Automation):用自然语言驱动浏览器完成重复网页任务

Browser Automation(@clawhub_peytoncasper/browser-automation,v1.0.1)由 peytoncasper 维护,基于 Stagehand CLI,用自然语言与 browser 子命令完成导航、点击、填表、截屏与数据提取。SkillHub 可通过 skillhub install 装入 Cursor 等 Agent;OpenClaw 可用 openclaw skills install。支持本地 Chrome 或配置 BROWSERBASE_API_KEY 的远程 Browserbase。适合将重复网页办公流程交给 Agent 编排,需注意 Chrome/密钥与页面变更带来的失败排查。

Read More
BrowserSkill:让 Agent 像人一样操作网页

BrowserSkill 是微信对话平台团队在 SkillHub 发布的腾讯开源浏览器自动化 Skill(@tencent-adm/browserskill,v1.0.7),面向 WorkBuddy 等 Agent,支持像人一样填表、点击、翻页、截图并完成复杂网页任务。本文基于 SkillHub 公开信息说明定位、能力与安装方式:通过 skillhub CLI 将技能装到对应 Agent 的 skills 目录,并给出 WorkBuddy、Cursor、Claude Code、Codex 等路径参考;未编造未公开的 API 与示例。

Read More
DeepSeek Harness 浏览器插件介绍

DeepSeek Harness 插件 `ricardochen1996/dsh-browser-use` 是一个浏览器自动化插件,利用 Jev Ultrafast 引擎将页面读取为索引化的动作空间,并实行单步受保护操作,降低了直接操作风险并提升了决策效率。插件采用代理架构,主对话仅提供管理接口,具体的浏览操作(如打开、点击、截图等)委托给子代理执行。该插件简化了浏览器集成复杂度,无需手动配置即可使用,适合需要稳定自动化能力的场景。

Read More
Agent Browser Operator:面向登录与会话的交互式浏览器 Skill

Agent Browser Operator 是 SkillHub 社区 Skill(v1.0.0,ai-agent),用于登录流程、动态页面和会话感知的交互式网站操作,标签 requires_api_key=false。文章说明其定位、三项核心能力、通过 skillhub CLI 安装到 Cursor/Claude/Codex 等 skills 目录的步骤,以及无官方端到端示例时如何按简介使用;并给出适用场景、环境限制与 SkillHub 官方链接。

Read More
Chrome调用:为 Agent 准备 Chrome CDP 运行时

Chrome调用(@user_2a19425c/chrome-cdp-runtime)是 SkillHub 社区 Skill,用 SKILL.md 约束 Agent 仅通过 Chrome CDP 与 9222 端口准备浏览器运行时:检查端口、接管已有实例或在提供可执行路径时启动调试 Chrome,并完成 tab 选择与基础页面操作。本文说明其核心能力、硬约束、可选输入、skillhub 安装方式及文档中的执行流程,适合需要与本机 Chrome 远程调试对齐的自动化场景。

Read More
Kimi WebBridge Pro:让 Agent 操作本机真实登录态浏览器

Kimi WebBridge Pro 是 SkillHub 社区 Skill(v1.0.1,@user_9ed681d6/kimi-webbridge-pro),面向本地 AI Agent,经本机 Kimi WebBridge 守护进程操控用户真实、已登录的浏览器,强调隐私最小化与标签页安全。适用于需要真实会话、已有标签页、截图、填表、确认上传、保存 PDF 或浏览器侧网络诊断的任务;纯搜索与无需浏览器状态的工作应改用 API、MCP 或检索方案。安装需 SkillHub CLI,并用 skillhub install 指定 Cursor 等客户端的 skills 目录;requires_api_key=false。

Read More
京东数据采集:用浏览器自动化从搜索页导出商品 CSV

介绍 SkillHub 社区 Skill「京东数据采集」(@user_f4e1eca6/jd-scrape,v1.0.0)。该 Skill 通过浏览器自动化从京东搜索页采集商品,支持多页翻页、懒加载滚动与 CSV 导出,无需 API Key。文章说明其定位、核心能力、基于 skillhub CLI 的安装方式(含 Cursor、Claude Code、Codex 的 skills 目录),以及使用注意与合规提醒,并给出官方页面链接。

Read More
dsh-doyourweb:复用现有 Chrome/Edge 标签页的 DeepSeek Harness 插件

`dsh-doyourweb` 是 DeepSeek Harness 的联网工具插件,旨在复用 Chrome/Edge 已登录标签页的会话状态。它支持通过菜单集成、紧凑阅读、批量操作及数据提取等功能,直接控制现有页面。安装需 Node.js 22+,通过本地扩展连接。该工具降低了自动化环境配置成本,无需重新登录即可处理网页任务。

Read More
browser-act:让 Agent 用真实浏览器完成页面任务

browser-act(@user_2cc35c71/browser-act,v1.0.4)是 SkillHub 社区发布的 ai-agent 类 Skill,将浏览器自动化封装为 CLI,供 Agent 提取渲染后页面、维持会话、处理验证、填表点击、截图及捕获 XHR/HAR 等。官方要求勿在 Bash 中直接执行 browser-act,应先调用 Skill,并优先于内置 fetch。本文基于 SkillHub 公开说明整理定位、能力、skillhub 安装路径与 browser list 示例,附官方地址与使用注意。

Read More
WebSculpt:带有记忆的浏览器自动化——安装与引导入口

WebSculpt 通过 CLI 记忆层把 Agent 首次跑通的浏览器操作固化为可复用命令,经 CDP 复用本机登录态。SkillHub 上的 @user_787620f7/websculpt(v1.0.3)是安装与引导入口:在 websculpt 命令不可用或需修复时,按 skillhub.md 安装该坐标;环境已有 websculpt-explore 等生命周期 Skill 时则应直接使用后者。文章说明四段式协作 Skill、典型 Explore/Maintain 示例及适用注意。

Read More
浏览器自动化脚本生成器:用自然语言产出 Playwright 风格 Python 脚本

浏览器自动化脚本生成器(@user_f28fdad8/oc-browser-auto)是 SkillHub 社区 dev-programming 类 Skill,v1.0.0,无需 API Key。它根据登录、导航、表单、爬虫等场景的操作描述,引导 Agent 输出基于 Playwright 语法模板的 Python(.py)脚本。安装可参照 skillhub.cn/install/skillhub.md,用 skillhub install 并 --dir 指向 Cursor 等客户端的 skills 目录。适合需要快速搭脚本又愿在本地补全参数与合规评估的开发者。

Read More
头条自动发布:用 Agent Skill 跑通头条号微头条与文章发布

SkillHub 社区 Skill「头条自动发布」(@user_8eb1b5b0/toutiao-auto-publish-new,v1.0.0,content-creation,无需 API Key)以 SKILL.md 规范头条号微头条与文章的选题、写作、配图与 browser/CDP 发布流程。本文介绍其核心能力、skillhub 安装方式、典型 browser 操作示例,以及登录态、页面 ref 与发布频率等注意事项;并说明需配合 browser-cdp 与 OpenClaw 类环境使用。

Read More
DeepSeek Harness 插件:dsh-plugin-cdp

`dsh-plugin-cdp` 是一款基于 Chrome DevTools Protocol 的 DeepSeek Harness 插件,旨在解决浏览器扩展自动化中标签页失效及弹窗不可信等问题。该插件不依赖外部库,通过 CDP 驱动受控 Chrome 实例,提供零运行时依赖的自动化方案。核心功能包括:通过 CDP 进行精确的页面操作与校验、首次使用时收集投递规则、生成纯文本可视化报告,以及支持无停机热替换机制,确保规则更新时任务不断档。插件提供 15 个工具支持导航、截图、数据读取及报告生成,开发者需在配置中指定 Chrome 路径与用户数据目录。

Read More
浏览器自动化:用目标驱动的浏览哲学约束 Agent 上网

SkillHub 社区 Skill「浏览器自动化」(@user_463e8e26/web-access-plus,v1.0.0)面向 ai-agent 场景,用四步框架约束 Agent 做网页任务:先定成功标准、再选起点(含需登录或静态难达时倾向 CDP)、过程用证据校验、达标即停。无需 API Key。文章说明其定位、核心逻辑,以及通过 skillhub.md 与 skillhub install 安装到 Cursor 等 skills 目录的方式,并提示其为方法论而非独立浏览器工具。

Read More
egobrowser:让人与 AI Agent 共用登录态的 Chromium 浏览器 Skill

egobrowser 是 SkillHub 社区 Skill(@user_57645b9d/ego-browser,v2.0.0),对接 citrolabs 的 ego-lite 浏览器。Agent 在独立 Space 中完成打开页面、填表、点击、截图与数据提取,并复用本机登录态,避免与日常浏览抢标签。文章说明核心能力、SkillHub 与 npx 安装方式、Cursor skills 目录示例,以及官方 /ego-browser 任务示例与平台限制。

Read More
@dsh-external/dsh-browser-use:DeepSeek Harness 的本机浏览器控制能力

DeepSeek Harness (DSH) 的 `@dsh-external/dsh-browser-use` 插件移植自 ZCode 的浏览器控制技能,核心采用 ARIA 快照优先的操作范式。该插件通过工具与可视化面板驱动本机 Chromium,支持网页打开、交互、快照读取及截图等功能,并依赖 Playwright 实现懒启动与持久化登录。安装后 Agent 可接管浏览器完成自动化任务,但需注意页面刷新后模块失效、截图需挂载服务、以及同源鉴权等限制。

Read More
浏览器自动化助手:把网页操作写成可执行的 Agent 指南

浏览器自动化助手(@user_c562f70c/browser-auto)是 SkillHub 社区 Skill,版本 1.0.0,面向 Agent 提供点击、表单、截图、多标签、登录、上传下载与数据抓取等流程指引,适用于 OpenClaw、Playwright、Selenium。文章说明其定位、能力与 SkillHub CLI 安装方式(含 Cursor、Claude Code 等 skills 目录),并给出安装后通过对话引用的典型起步步骤。无需 API Key。

Read More
BrowserSkill 浏览器自启动器:让 bsk 抓取任务自己拉起 Edge

BrowserSkill(bsk)抓取与自动化要求本机 Edge 已开且扩展连上 daemon,无人值守任务常因 browsers connected 0 中断。SkillHub 社区 Skill「BrowserSkill 浏览器自启动器」(@user_656c2e70/bsk-browser-launcher,v1.0.0)通过 launch_browser.py 做 bsk status 自检、按需启动 Edge、轮询等待连接,并以退出码 0/3/4 约定就绪、超时与建会话失败;可选 --start-session 输出 SESSION=id。文章说明安装命令、典型 bsk 会话骨架及 Windows 路径依赖。

Read More
QQBrowserUse:面向 AI Agent 的浏览器自动化 CLI

QQBrowserUse(@tencent-adm/qqbrowser-skill v1.0.16)由腾讯 QQ 浏览器团队发布,为 AI Agent 提供 qqbrowser-skill CLI:在独立 Tab Group 中完成导航、填表、截图、下载、页面状态读取与结构化抽取,并支持在用户明确要求时录制任务及 Playbook 回放。本文说明 Skill 触发条件、核心命令、SkillHub 与 pipx/pip 安装步骤、Quick Start 示例,以及平台支持与 tabId、会话隔离等注意事项;不要求 API Key。

Read More
浏览器自动化工具箱:难抓站点的多方案降级 Skill

浏览器自动化工具箱(Browser Automation Toolbox)是 SkillHub 社区 Skill,坐标 @user_7495c5ed/browser-automation-toolbox,v2.0.11,分类 data-analysis。它提供 CloakBrowser、browser-act、Kimi WebBridge、Playwright 四套爬虫自动降级,以及 BrowserSkill 互补自动化,面向反爬难站、SPA 滚动、登录与验证码人工接管、AI 平台自动化等场景。安装需 SkillHub CLI,使用 skillhub install 并 --dir 指向 Agent 的 skills 目录(如 Cursor 的 ~/.cursor/skills/)。requires_api_key=false。

Read More
Remote noVNC + Chromium CDP: Visual Browser Automation Shared with an Agent

Introduces the SkillHub community Skill @user_f72e84b4/tencent-novnc-chromium-cdp (v1.0.44). On Linux, it combines noVNC, Chromium, and CDP to provide a remote visual browser. On Windows, it can take over local Edge/Chrome CDP, allowing the agent and the operator to share the same browser instance, with login and slider CAPTCHAs manually handled in the UI as a fallback. It also explains installing the skillhub CLI, pointing --dir to each agent’s skills directory, and accessing the official page, without inventing undisclosed deployment parameters.

Read More
Agent Browser: A Browser Automation CLI for AI Agents

Agent Browser is an AI-agent Skill maintained by rez0. It enables agents such as Cursor to drive browsers via Chrome CDP for browsing, form filling, clicking, screenshots, and data extraction, using an npm global CLI and SkillHub/OpenClaw skill packages. Key features include accessibility tree snapshots with @eN element references, a native Rust CLI, no dependency on Playwright or Puppeteer, and the ability to load extensions such as Electron, Slack, and QA via `skills get`. This article summarizes installation paths and a typical getting-started workflow based on the official SkillHub and ClawHub pages.

Read More
firecrawl-build-interact: Assembling Multi-Step Firecrawl /interact Browser Workflows in Business Code

firecrawl-build-interact is an official Firecrawl Agent Skill used to assemble Firecrawl /interact workflows in business code. When /scrape cannot handle logins, pagination, form filling, infinite scrolling, or JavaScript single-page applications, it encapsulates patterns such as steps, waits, and authenticated navigation, allowing multi-step browser operations to be treated as a single call. The installation command is npx skills add with the target repository and skill name. It is suitable for scenarios such as authenticated scraping, pagination collection, and multi-step form processing.

Read More
DeepSeek Harness Chrome Browser Extension

The `dsh-chrome-browser` plugin for DeepSeek Harness (DSH) directly drives a real browser via the Chrome DevTools Protocol (CDP), providing the model with internet access capabilities. It does not require frequent permission approvals and supports page navigation, screenshots, DOM manipulation, JavaScript execution, and interactive control, while not being subject to sandbox restrictions. Installation requires packaging via npm and loading it into the Harness process, with support for Linux, macOS, and Windows. When using it, be sure to install CJK fonts and account for platform-specific permission differences.

Read More
dsh-playwright: Official Playwright MCP Integration, Structured Accessibility Tree Browser Automation

dsh-playwright is a profile bundle for DeepSeek Harness that integrates the official Microsoft Playwright MCP, enabling agents to control real browsers directly through structured accessibility trees without relying on vision models. The plugin supports navigation, clicking, text entry, and screenshots, and provides tab management and request interception capabilities. It requires environment prerequisites such as Node.js and can be installed and enabled with a single command. The plugin runs under the current user's permissions, does not use API keys or telemetry, and is part of DeepSeek Harness's "everything is a plugin" ecosystem.

Read More
DSH Browser: A Headless Browser Automation Plugin for DeepSeek Harness

dsh-browser is a browser automation plugin for DeepSeek Harness, designed to address the session confusion and security risks caused by traditional drivers. Built on Playwright 1.62.1, it provides headless-first, fully isolated Chromium sessions and a semantic action interface. Through bounded evidence and deterministic checks, it ensures that Agent actions are safe and traceable. The plugin supports a standard workflow: installing and starting a session, observing a page, performing actions, and ending the session, and is suitable for web testing and data scraping scenarios. The current version is 0.1.0-rc1, a pre-release.

Read More
dsh-browser-use

`dsh-browser-use` is a browser automation plugin for DeepSeek Harness that connects browser control capabilities through the Open Browser Use WebMCP ecosystem. The plugin requires Node.js 22.19 or later; after installation, it connects to the MCP Server and exposes tools under the `mcp__open_browser_use__*` namespace. By design, it uses Host Root isolation to avoid permission exposure, and the child process of a given instance supports only a single Session. It does not reimplement WebMCP and is only responsible for bridging, with the security boundary managed by the Harness runtime.

Read More
dsh-browser-playwright-codex: Enabling AI to Operate in Real Browsers

`dsh-browser-playwright-codex` is a browser automation plugin for DeepSeek Harness (DSH) that uses Playwright to drive AI-based web operations in a visible browser window. Its core features include reusing logged-in accounts, preserving browser state, supporting multi-tab management and screenshots, and enabling stable operations through `ref` references. When it encounters CAPTCHAs or pop-up dialogs, the plugin automatically pauses and waits for human intervention, thereby addressing real-world login and interaction challenges. After installation, high-risk `evaluate` and `cdp` tools are disabled by default, and users must first obtain a Snapshot before performing subsequent actions. It is suitable for use cases such as list scraping, form filling, and order querying.

Read More
dsh-chrome-operation: Turning Local Chrome into a DSH Toolkit

`dsh-chrome-operation` is a client plugin for DeepSeek Harness (DSH) that connects to your local Chrome browser and converts browser operations into tools that DSH can recognize. It communicates with the installed `chrome-operation` extension over WebSocket. Installation requires two steps: first load the `chrome-operation` extension in `chrome://extensions`, then run `dsh plugin add dsh-chrome-operation` in your terminal and restart DSH. The plugin supports various browser operations, including snapshots, navigation, clicks, input, scrolling, tab management, and history. It only listens on the local loopback address.

Read More
dsh-browser: Providing a Local Browser for Each DeepSeek Harness Session

`dsh-browser` is a DeepSeek Harness plugin that provides a dedicated, real browser environment for each conversation session. It mirrors the browser view to the sidebar via the CDP port, allowing agents to drive the browser with nine tool sets (such as page snapshots, element lookup, and code execution). It supports real mouse and text event input and enables developers to debug via DevTools or Playwright, making it suitable for scenarios requiring in-depth debugging of an agent's web interaction behavior.

Read More
dsh-plugin-browser: A Native Browser Automation Plugin for DeepSeek Harness

`dsh-plugin-browser` is a native browser automation agent plugin for DeepSeek Harness (DSH), maintained by menotbobbybrown. It addresses the challenges agents face when handling dynamic web pages, performing structured scraping, and conducting visual verification. Its core features include web page navigation, semantic Markdown extraction, high-resolution screenshots, and CLI tools. Once installed, agents can automatically invoke tools such as `browser_navigate` and `browser_extract` through Cordis Native, while developers can also interact with web pages directly via the command line, making it well suited for scenarios that require advanced web capabilities.

Read More
dsh-browser-annotate: A Browser Annotation Tool Integrated with Real Chromium

`dsh-browser-annotate` is a browser interaction plugin for DeepSeek Harness (DSH), designed to address the lack of intuitive feedback and advanced debugging capabilities in agents. It provides a real Chromium environment and CDP access through an Electron desktop app, supporting 12 operational tools, including navigation, screenshots, clicking, dragging, popup handling, and custom code execution. A key feature is its right-click annotation mechanism, which lets users enqueue element information—such as selectors, screenshots, and styles—and inject it into the conversation, with a dual-pane layout and local communication support. Installation requires DSH plugin commands. Known limitations include the inability to close tabs automatically and the lack of support for creating new browser contexts.

Read More
DeepSeek Harness Plugin: dsh-tmwebdriver - Run Any JavaScript in a Real Browser

`dsh-tmwebdriver` is a plugin in the DSH ecosystem that overcomes browser operation restrictions. By integrating TMWebDriver, it enables agents to execute arbitrary JavaScript in a real, logged-in Chrome browser tab. Core features include `browser_execute_js` (covering DevTools-level operations), page interaction, screenshots, and cookie management. The plugin can effectively bypass anti-scraping measures and supports adjusting parameters such as timeouts via configuration files, making it ideal for scenarios that involve handling CAPTCHAs, maintaining logged-in sessions, or interacting with complex JavaScript applications.

Read More
dsh-browser-use: DeepSeek Harness Browser Automation Plugin

`dsh-browser-use` is a browser automation plugin for DeepSeek Harness (DSH), maintained by `helloproblems`. It uses a three-layer architecture consisting of Hub, Backend, and Domain, supports Chrome and Edge backends, and provides an Agent isolation mechanism. The plugin achieves browser control through `chrome-devtools-mcp` and Playwright MCP, supports hot backend switching, and is suitable for web data scraping, automated testing, and interaction simulation. Installation requires Node.js ^22.19.0 or >=24.0.0 and a compatible DSH environment.

Read More
Browser MCP: A Local MCP Server for Accessing Authenticated Content via Real Chrome

Browser MCP solves the challenge of AI accessing login-required websites such as Zhihu and Xiaohongshu. It uses a local Chrome extension to control a real browser, leveraging existing logged-in sessions to interact with web pages, without reverse-engineering APIs or handling cookies. It supports searching, scraping, downloading, and background automation, keeps all data stored locally, and ensures compliance and security. After installation, AI can interact with web pages directly in conversations using natural-language instructions.

Read More
DeepSeek Harness (DSH) Browser Automation Plugin: dsh-playwright

The `dsh-playwright` plugin for DeepSeek Harness (DSH) leverages Playwright to provide agents with a set of browser automation tools that are dynamically loaded at runtime. It supports both self-hosted and mounted modes, offering complete capabilities—from opening pages and click-based interactions to data scraping and capturing screenshots—and runs directly in the host process without requiring an external bridge service. Installation requires manually downloading Chromium, and it supports switching between Chinese and English as well as fast error propagation. It is currently compatible only with DSH 0.2.0-rc.2.

Read More
dsh-webops-plugin: DeepSeek Harness Web Operations and Debugging Plugin

This plugin is designed to provide DeepSeek Harness with comprehensive web debugging and manipulation capabilities. It addresses the cumbersome nature of browser environment operations through standalone windows, multi-tab management, and CDP/Electron support. Core features include element manipulation (clicking, filling, and scrolling), accessibility snapshots, console and network data collection, and advanced CDP command execution. The plugin supports portable desktop installation, employs a tab-exclusive mechanism, and offers a manual takeover mode. It provides agent developers with a fine-grained tool for controlling browser behavior.

Read More
dsh-browser-ops: DeepSeek Harness Browser Automation Plugin

`dsh-browser-ops` is a DeepSeek Harness plugin designed to address the limitation that agents cannot operate browsers directly. It provides 27 `browser_*` tools that allow agents to take control of a Chrome browser and perform operations such as taking snapshots, clicking, and filling forms. Its core features include recording manual workflows into JSON with variable slots, supporting parameterized replay to generate test data, and offering a Web GUI panel along with two browser modes: Managed and Connect. It is suitable for scenarios requiring automated browser interaction, reuse of manual workflows, and test data generation, but it does not currently support file uploads, drag-and-drop, or cross-tab operations.

Read More
@kasenri/dsh-browser: DeepSeek Harness Browser Automation Plugin

The DeepSeek Harness plugin `@kasenri/dsh-browser` is a community plugin designed to address the limitation that AI agents cannot directly operate a browser. It injects browser automation capabilities through the native Harness infrastructure, supporting tasks such as web testing and information collection. The plugin depends on specific versions of DSH and `agent-browser`, and provides core functions such as opening web pages, clicking, typing, and reading content, while supporting multiple input modes. To ensure security, it includes built-in protection mechanisms to prevent stale references, tab drift, and sensitive information leakage, validate disk artifacts, restrict domains, and enforce manual verification. During installation, ensure that the required dependent components are available in PATH and restart the profile for the changes to take effect.

Read More
dsh-opencli: Connecting Real Browsers and Site Adapters to DeepSeek Harness

The `dsh-opencli` plugin for DSH Harness provides “task execution” capabilities via OpenCLI. It drives actual browser windows through a browser extension and supports 176 sites with 200+ commands. Users can retrieve structured data and perform actions such as posting or approving, with all operations requiring approval by default. The plugin supports scheduled tasks, auditing, and concurrent multi-site execution. Installation requires deploying the OpenCLI environment first; ensure Windows paths are configured correctly, and note that headless servers are not supported.

Read More
BrowserRig: A Local Driver Plugin for Reusing Logged-In Chromium Sessions

BrowserRig is an open-source local browser automation plugin designed to address the pain point of reusing existing browser sessions in agent development. It connects to an already logged-in Chromium instance through a Chrome extension, reusing cookies, extensions, and login state without requiring the remote debugging port to be enabled. Its core features include session takeover, background execution, and integration with tools such as DeepSeek Harness. This solution avoids repeated popup confirmations and enables code-level execution control, making it particularly well suited for development and testing scenarios that require persistent login sessions or frequent window switching.

Read More
dsh-pc-pilot: A Windows Desktop and Browser Automation Plugin

`dsh-pc-pilot` is a DeepSeek Harness plugin running on Windows 10/11 that provides agents with integrated local desktop and browser operation capabilities. By leveraging the UIA accessibility tree, background synthetic input, and persistent Chromium CDP sessions, it addresses the issues of traditional automation stealing focus and obscuring windows. The plugin supports application and window management, basic interactions, and browser navigation, with advanced features such as identity recognition, state restoration, and batch execution. It runs with the current process permissions, does not support virtual machines or hidden sessions, and is designed to provide a secure and stable foundation for agent environment operations.

Read More
dsh-tui-browser-use: A Browser Automation Sub-Plugin for DeepSeek Harness

`dsh-tui-browser-use` is a browser automation plugin designed for DeepSeek Harness (DSH), intended to give Agents the ability to view and interact with web pages. It is driven by Playwright, provides 21 `browser_*` tools, and supports navigation, interaction, data extraction, and session management. A key feature is its native support for DeepSeek vision models, with automatic screenshot segmentation, security protection, and persistent login sessions. It depends on `dsh-tui` v0.10.0 and is currently archived.

Read More
DeepSeek Harness Plugin: dsh-web-search-opencli

`dsh-web-search-opencli` is a DeepSeek Harness plugin designed to address the instability of traditional web scraping under anti-crawler mechanisms. It uses OpenCLI to directly control Chrome or Edge browsers and invokes Google AI Mode to perform searches. The plugin extracts Google AI's answers and sources into a standard `WebSearchResult`, supporting Markdown conversion, session reuse, and highly configurable behavior. After installation, DSH automatically applies the configuration; when an agent calls `web_search`, the plugin drives the browser, waits for the AI response, and returns the result. Before use, Node.js, DSH, and OpenCLI must be installed, and AI Mode support is required.

Read More
lexmount/dsh-browser: DeepSeek Harness Browser Automation Tool

`lexmount/dsh-browser` is a lightweight plugin for DeepSeek Harness (DSH) that injects Lexmount cloud browser capabilities into DSH by registering native tools, enabling models to call browser operations directly. The plugin provides 31 native tools covering page diagnostics, authentication, browser operations, and advanced CDP capabilities. Installation requires compatible Node.js and DSH versions and supports Windows, macOS, and Linux platforms. It also features automatic downloading and caching of `browser-cli`, strong security with injection prevention, and offline reuse. Once installed, browser automation can be achieved in DSH through steps such as initialization, authentication, session management, and page operations.

Read More
Web Bridge — DSH Plugin: Real Browser Automation

Web Bridge is an MCP server plugin built on real browsers (Chromium/Firefox/WebKit) that provides agents with precise page manipulation capabilities through an accessibility tree snapshot mechanism. It supports navigation, clicking, form filling, screenshots, and script execution, and features tab management, multiple wait types, and session persistence. This tool is suitable for scenarios requiring high-precision interaction and dynamic web page processing, and can be deployed flexibly via configuration files or environment variables.

Read More
DeepSeek Harness Browser Automation Plugin: Security-First Web Interaction Capabilities

dsh-plugin-browser-use is a browser automation plugin for DeepSeek Harness (DSH). Built on Chromium and Playwright, it provides 15 core tools, including navigation, page interaction, and screenshots. Its core advantage is a security-first design: policies are re-evaluated after each action, strictly restricting file paths and network access while supporting allowlists, protocol restrictions, and private network control. Installation requires Node.js 22.19+ or 24+, and Chromium must be configured. This plugin provides DSH agents with secure and controllable web interaction capabilities.

Read More
dsh-computer-use: General-Purpose Computer Use Capabilities of DeepSeek Harness

`dsh-computer-use` is a plugin for DeepSeek Harness that provides agents with a unified computer operation interface. It wraps the underlying drivers through `ctx.computerUse`, supports multiple backends such as Playwright and the Windows native assistant, and integrates accessibility and visual perception modes. The plugin provides security mechanisms, Cookie Bridge support, and depends on several DSH base plugins. Its configuration is flexible and can be modified in real time via the GUI, but there are platform limitations and configuration isolation issues caused by App-Bound Encryption. When using it, be careful to import only required data and configure a domain whitelist.

Read More
dsh-pilot: A Browser Automation Extension Based on DSH Session Permissions

`dsh-pilot` is a DSH plugin that intercepts requests at the browser network layer to enforce a DSH session-based permission model and prevent page navigation from bypassing checks. It provides tools for navigation, snapshots, interaction, waiting, screenshots, and closing tabs. The plugin's permissions are based on local access, pre-authorized origins, and session policies, and password field input is denied by default. Known limitations include the maintenance cost of the Ref mechanism, headless rendering discrepancies, and the risk of name conflicts.

Read More
dsh-cdp-browser: CDP Browser Automation with Zero Process Startup

`zaiwenj/dsh-cdp-browser` is a plugin for DeepSeek Harness designed to address the limitations of launching browsers in sandboxed environments and the high cost of conventional visual verification. The plugin controls an already running local browser via the CDP protocol, using Node’s `fetch` and `WebSocket` to achieve subprocess-free automation and pixel-level checks without relying on a vision model. It supports page operations, script execution, and multiple assertion types (pixel, CSS, DOM), but page reloading should be handled carefully to avoid state interference, and sampling at gradient stop points is recommended.

Read More
ylwl1997/dsh-browser: Real Browser Automation Based on bsk

dsh-browser is a plugin for DeepSeek Harness that addresses the challenge of agents accessing real web content and interacting with live web pages by introducing bsk. It enables agents to navigate, click, fill out forms, and take screenshots like humans. The plugin offers core features including navigation, reading the accessibility tree, clicking, form filling, screenshot capture, and semantic observation. Before use, ensure that the system environment variable PATH includes bsk and that the browser skill extension is loaded in DSH. Released under the MIT license, this tool is designed to meet the need for agents to perform complex web interaction tasks.

Read More
dsh-live-loop: A Frontend Runtime Verification Plugin for DeepSeek Harness

`dsh-live-loop` is a frontend runtime validation plugin for DeepSeek Harness (DSH), designed to ensure that web applications load correctly and behave as expected by performing detection, interaction, visual comparison, and evidence generation. It does not modify code and is responsible only for the validation process, supporting frameworks such as Vite, React, and Vue while providing isolated environments, bounded observation, and a strict four-state verdict mechanism. It requires DSH version 0.1.0-rc.7 and Node.js 22.19.0+, giving agents a closed-loop capability from detection and remediation to validation to ensure rigor in frontend automation workflows.

Read More
dsh-browser-bridge: Control/Read Browsers via Local WebSocket

The DeepSeek Harness plugin `dsh-browser-bridge` enables agents to control real browsers via a local WebSocket. It provides 13 model-side tools covering status checks, page reading, interactive operations, and code execution. Installation requires the server-side plugin and a browser extension. It binds to a local address by default and should be used only in trusted environments.

Read More
dsh-tool-playwright: Enabling AI to See and Interact with Web Pages in the Browser

dsh-tool-playwright is a native port of Microsoft Playwright. As a plugin for DeepSeek Harness, it gives AI direct control over a real browser, enabling it to operate web pages, fill out forms, and scrape data just like a human. The plugin does not require a separate server and runs directly inside Harness. To install it, clone the codebase and register it using the link method, then restart Harness to make it take effect. It provides page navigation, interaction, screenshots, multi-tab management, and advanced debugging features, and must be configured via cordis.patch.yml. This tool integrates browser automation capabilities into Harness and simplifies the process of building agents with internet-interaction capabilities.

Read More
dsh-cdp-live-view: DSH Sidebar CDP Live View Plugin

dsh-cdp-live-view is a DeepSeek Harness (DSH) plugin that uses the Chrome DevTools Protocol (CDP) to render remote or headless Chromium browser sessions in real time in a Web GUI sidebar. It supports viewing page pixel streams, managing targets, navigating, and controlled mouse/keyboard interaction, with interaction gated by a dual-switch control. It aims to provide a secure remote visualization solution, suitable for debugging automation workflows or as a lightweight VNC alternative. Installation requires DSH ≥20 and `dsh-better-sidebar`.

Read More
dsh-cloudflare-browser-run: Provide real browser crawling capabilities for DeepSeek Harness

dsh-cloudflare-browser-run is a DSH plugin maintained by RealAlexandreAI that integrates Cloudflare Browser Run's headless Chrome into DSH, addressing the limitation of standard HTTP scraping in handling JavaScript-rendered pages. The plugin provides three tools: browse (returns clean markdown by default, with optional screenshots or PDF), screenshot, and pdf. Its design features include real browser access and SSRF protection.

Read More
dsh-browser-plus: A DSH Plugin Making Agent Browser Truly Visible and Controllable

## Introduction A common practice in browser automation is to run the Agent in a headless process invisible to the user. The problem is that the user has no way to confirm what is happening on the page or where the Agent clicked; when encountering steps that require login or manual judgment, there is also no natural handover method. dsh-browser-plus is designed to address this issue. It keeps the browser window visible to the user while providing the Agent with reliable Chrome DevTools Protocol (CDP) capabilities. Users can interact directly on the real page, and the Agent can execute, take over, and resume tasks synchronously. ## What is it dsh-browser-plus is a visual Agent styled with EGO, specifically crafted for DeepSeek Harness (DSH).

Read More
dsh-browser-control: A CDP plugin that lets DSH agents reuse your daily Chrome login state

## Introduction Connecting browsers to AI agents often involves launching a fresh headless instance. The problem lies in maintaining login sessions: many pages require logged-in cookies to display real content, and re-authenticating in isolated environments every time is costly. However, directly enabling `--remote-debugging-port` on a regular Chrome instance is explicitly prohibited in newer versions of Chrome—the default user-data-dir does not allow the debug port to be opened. Below, I introduce `dsh-browser-control`, a DeepSeek Harness (DSH) plugin that uses the Chrome DevTools Protocol (CDP) to drive the browser. Its core selling point is...

Read More
dsh-mindseye: Adding Image Understanding, Image Generation, and Visible Browser Automation to DeepSeek Harness

## Introduction In agent development, these issues frequently arise: the model is text-only, so when a user pastes a screenshot and asks, "What is this error?", the model cannot answer. To generate an image, one must manually open another tool, download the file, and drag it back into the session. To operate a webpage, mainstream solutions either run invisible headless browsers or directly take over the user's browser profile, which carries significant risks. dsh-mindseye (MindsEye) is specifically designed to address these three gaps. It is a plugin for DeepSeek Harness (DSH); in a nutshell, it provides intent-driven vision, image generation, and visible browser automation for

Read More
dsh-jiey-browser: Enable DeepSeek Harness to use real Jiey Browser via MCP

**Introduction** When integrating browser capabilities into DeepSeek Harness (DSH) agents, the typical approach is Playwright: a headless environment with a separate automation profile. The login state and cookies are completely isolated from the user's daily browser, and installation requires running `npx playwright install` first. If a task requires reusing sites that are already logged in, this path involves a lot of extra work. dsh-jiey-browser takes a different approach: it hands over the real Jiey Chromium (i.e., the user's own browser) to the agent via MCP, rather than spawning a separate headless environment. The following sections introduce what this plugin can do, how to install it, and how to use it. # What is this? dsh-ji...

Read More
dsh-playwright: Share a browser page with the model in DeepSeek Harness

dsh-playwright is an MIT-licensed DSH plugin maintained by JeremyGuo, allowing users and the model to share the same Playwright browser page. The model navigates, clicks, inputs text, and performs keyboard operations via browser_* tools, while users can monitor the process in real-time on the Browser Use panel and take over control at any time. Navigation.

Read More
dsh-plugin-chrome: Give DSH sessions a visible Chrome window

dsh-plugin-chrome is a browser visualization plugin for DeepSeek Harness (DSH) (MIT open source). Unlike headless black-box solutions, it launches a real, visible Chrome window for each session; the screen is streamed to the Web GUI in real-time via screencast, allowing users to observe at any time.

Read More
dsh-chrome: DeepSeek Harness Chrome Sidebar Plugin

dsh-chrome is a browser companion extension for DeepSeek Harness (DSH), designed to embed the full DSH web UI via the Chrome sidebar, granting local agents the ability to read pages, capture traffic, and control the browser. The plugin runs in a trusted local environment and requires Chrome 118+ and Node 18+. Core functions include automatically detecting the URL and content of the active tab (supporting SPA routes), capturing HTTP request details on demand (automatically masking sensitive headers and retaining the most recent 500 records), and providing basic control tools such as navigation, clicking, and opening new tabs. To ensure security, the plugin employs best-effort protection mechanisms, such as marking injected pages as untrusted data and utilizing intent-unlock.

Read More
dsh-plugin-browser: A tool for DSH to access a logged-in Chrome browser

dsh-plugin-browser is a browser automation plugin for DeepSeek Harness (DSH) based on Playwright and CDP technology, designed to address the issues of models accessing logged-in websites and sharing login states across multiple clients. The plugin registers 15 `browser_*` tools with DSH, supporting connection to a user's logged-in Chrome browser via the `real` mode or the launch of independent incognito sessions via the `isolated` mode. Core features include session management, ref-based page interactions (navigation, click, input, hover), handling of tabs and popups, JavaScript execution, reading storage data, screenshots, and waiting for human verification. Installation requires adding the plugin via the DSH CLI; if used...

Read More
dsh-web-auto-open: Automatically open browser when running dsh web

dsh-web-auto-open is a one-click Web launcher plugin for DSH, designed to resolve the issue of having to manually open the browser after starting `dsh web`. When starting the service, this plugin automatically invokes the system's default browser (supporting Windows, macOS, and Linux) to access the web interface, and it correctly handles the scenario where the system assigns a random real port when the port is set to 0, ensuring the correct address is opened. Installation is simple; it takes effect automatically as a profile bundle after executing `dsh plugin add dsh-web-auto-open`, requiring no additional configuration. Users can disable the auto-open functionality by setting `autoOpen: false` in the configuration file, or set `delayMs` to implement a delayed opening. This plugin is...

Read More
DSH-BrowserUse: A browser-use Style Browser Automation Plugin for DeepSeek Harness

`dsh-browseruse` is a browser automation plugin for DeepSeek Harness (DSH), implemented purely in JavaScript with no build step required. It drives a standalone Google Chrome instance via `playwright-core`, enabling operations such as opening web pages, reading content, clicking, inputting text, scrolling, and switching between tabs, while supporting autonomous decision-making based on page state during multi-step tasks. Core features include: 11 fine-grained `browser_*` tools; `browser_task` for natural language autonomous tasks (default maximum of 30 steps); `browser_schedule` for scheduled tasks within the current session; a user confirmation mechanism before sensitive operations (e.g., payments, deletions); manual takeover during CAPTCHA detection; and an independent Pr

Read More
dsh-browser-playwright: A Playwright-Based Browser Plugin for DeepSeek Harness

dsh-browser-playwright is an MIT-licensed browser plugin designed for DeepSeek Harness, aimed at addressing issues such as CSS selector mismatches or oversized DOM contexts in traditional solutions. By leveraging Playwright to drive a real browser, the plugin utilizes accessibility snapshots and stable element references (ref) to constrain the model's interaction interface, avoiding direct selector guessing or outputting the complete DOM. Its core features include support for tab management, screenshots, structured extraction, and controlled JavaScript execution, along with 17 tool interfaces prefixed with `browser_`. The plugin can automatically detect browser channels such as Chromium and Chrome, and supports restricting navigation domains.

Read More
DSH-Browser: A Playwright Browser Automation Plugin for DeepSeek Harness

dsh-browser is a Playwright-based browser automation plugin for DeepSeek Harness, designed to enable Agents to directly drive real web pages. It provides tools such as `browser_open`, `click`, `type`, `screenshot`, and `eval`, supporting page navigation, form interaction, JavaScript execution, content reading, and screenshot review. Browser state is maintained across tool calls, and it runs in headless mode by default. Installation requires `playwright-core` and a Chromium-based browser (it automatically attempts to use chromium/chrome/msedge; if missing, you can run `npx playwri

Read More
DSH-Cloak-Browser: Native Browser Plugin for DeepSeek Harness's CloakBrowser

`dsh-cloak-browser` is a community-maintained plugin for DeepSeek Harness (DSH), designed to integrate CloakBrowser capabilities into the Agent tool pipeline. It ensures session security by maintaining an isolated `BrowserContext` for each Agent, and returns restricted page snapshots along with short-lived element references (e.g., `p1:s3:e8`) after interactions such as navigation, clicking, input, and screenshots, enabling the Agent to continue operations based on these references without retaining extensive DOM data. Core features include on-demand Chromium startup, automatic license detection (supporting direct startup with Free keys and serializing concurrent sessions), proxy-aware GeoIP consistency, domain allowlist/blocklist rules, and handling for images and text—

Read More
dsh-pilot: Driving Local Browsers in DeepSeek Harness Chat

dsh-pilot is a plugin for the DeepSeek Harness Web GUI, designed to drive native Edge or Chrome browsers via CDP to perform browser operation tasks. It supports opening pages, reading structured text, clicking and inputting based on snapshot references, filling forms, and uploading files, all without requiring an API key or runtime dependencies. Users can observe real-time screenshots, URLs, status, and operation logs through the cockpit panel. Each agent session uses an isolated browser instance, ensuring that parallel sessions do not interfere with each other. Installation requires Node >= 22, and the plugin can be added via the `dsh plugin` command from GitHub or a local directory, followed by restarting the `dsh web` process. Key limitations include support for only headless mode, single-session single-tab functionality, and restricted interaction capabilities.

Read More
dsh-playwright-browser: Providing Playwright Multi-tab Browser Automation for DSH Agents

dsh-playwright-browser is a DSH networking tool plugin maintained by Clizo1209, based on Playwright, which registers ten browser_* tools for intelligent agents. It supports multi-tab, semantic positioning, bounded snapshots, and screenshots. The article covers installation commands, cordis.patch.yml configuration, typical invocation flow, and security considerations. The plugin runs with DSH process permissions; please verify the source code and MIT license before installation.

Read More
dsh-reef: One-Time Installation for Browser, MCP, and Code Hosting Integration

dsh-reef is a DSH plugin bundle maintained by huey1in, offering five modules with a single installation: native embedded panels, 21 browser_* browser tools, MCP Server, GitHub/GitLab read-only tools, and webhook automatic reviews. It supports multi-tab, multi-profile browser sessions, exposing DSH sessions to external MCP clients, automatic PR/MR reviews, and issue auto-fixing loops. The installation command is dsh plugin --profile web add dsh-reef, licensed under MIT, and requires Node.js >= 22.

Read More
dsh-browser: A Plugin for Providing Shared Real Browsers to DSH

dsh-builtin-browser (wqty123/dsh-browser) is a client-side plugin for DSH that provides agents with a real, visible browser that can be shared with users. Upon installation, it offers 20 browser_* tools, supporting DOM-level operations, multi-tab management, task-level session isolation, persistent login state, human verification detection, and cookie-enabled downloads. When a desktop shell is available, it embeds the view; in a pure DSH web environment, it can self-host and launch an Electron window. Installation: dsh plugin --profile web add dsh-builtin-browser. Licensed under MIT, currently at version 0.1.15.

Read More
Connect Playwright Browsers to DeepSeek Harness with dsh-playwright-browser

dsh-playwright-browser is a community plugin for DeepSeek Harness maintained by Clizo1209. It provides 10 browser_* tools, semantic positioning and multi-tab control for agents via Playwright. This article verifies the installation commands, configuration items and security boundaries against the directory page, GitHub README and npm version 0.1.3: no page eval is provided, and the plugin runs with the permissions of the current dsh process.

Read More
Integrate ego-lite Browser into DeepSeek Harness using ego-browser

Ego-browser is a DeepSeek Harness community plugin maintained by Fisfzy, which integrates the Chromium framework for AI Agents, CitroLabs/ego-lite, into Harness. The current version of this repository is 0.8.0, and its source code registers 32 ego_* tools including semantic snapshots, click-to-fill forms, screenshots, CDP, task space isolation and more, with built-in ego runtime and real-time observation window. The community directory installation command is `dsh plugin add github:Fisfzy/ego-browser`. This plugin has no affiliation with DeepSeek/HuanFang; the plugin runs with the permissions of the current dsh process, and you should check the source code and license before installation.

Read More
Use dsh-plugin to drive the local Tabbit browser with DeepSeek Harness

The dsh-plugin maintained by Tabbit-Browser integrates the CLI and Runtime of the Tabbit Browser into DeepSeek Harness. After installation, the Tabbit Browser Skill will be automatically loaded, and `tabbit_browser_install` will be used to detect the 1.9.0 official version and download and install the installation package according to the region when necessary. This article verifies the installation commands, task space usage, Windows permission boundaries, and restrictions such as unavailability of remote sandboxes based on the directory page, GitHub README and source code.

Read More
Drive Real Browsers via Terminal: Getting Started with OpenAI's Official Playwright Skill

OpenAI's featured Skill "playwright" wraps the Playwright Agent CLI into a terminal browser automation workflow: opening pages, taking snapshots to obtain element references, filling forms and clicking, taking screenshots, and debugging with Trace. Verified based on the official SKILL.md, this article introduces the installation methods (Codex skill-installer, Cursor .cursor/skills), the usage of the wrapper script, typical command examples, as well as precautions such as re-snapshotting when refs fail and not writing @playwright/test by default. It is suitable for E2E exploratory automation and UI process troubleshooting scenarios.

Read More