[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-74273":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":39,"readmeContent":40,"aiSummary":41,"trendingCount":16,"starSnapshotCount":16,"syncStatus":15,"lastSyncTime":42,"discoverSource":43},74273,"skills","browser-act\u002Fskills","browser-act","Browser automation CLI built for AI agents. Break through anti-bot walls, hand off to humans across platforms when stuck. Parallel multi-task execution, independent multi-session operation, isolated multi-account browsing.","https:\u002F\u002Fwww.browseract.com\u002F",null,"Python",2334,90,6,2,0,301,813,1137,903,27.88,"MIT License",false,"main",true,[27,28,29,30,31,32,33,34,35,36,37,38],"ai-agents","automation","claude-code","claude-code-skills","cursor","data-extraction","no-code","openclaw","openclaw-skills","web-data-extraction","web-scraping","web-scraping-api","2026-06-12 02:03:24","\u003Cdiv align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fwww.browseract.com\" style=\"text-decoration: none;\">\n    \u003Cimg src=\"https:\u002F\u002Fbrowseract-prod.browseract.com\u002Fprod\u002Ftools\u002F20260205-154549.png\" alt=\"BrowserAct Logo\" width=\"150\">\n  \u003C\u002Fa>\n  \u003Ch1>BrowserAct Skills\u003C\u002Fh1>\n\n  \u003Cp>\n    \u003Ca href=\"https:\u002F\u002Fdiscord.com\u002Finvite\u002FUpnCKd7GaU\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fdiscord\u002F1234567890?label=Discord&logo=discord&color=7289DA\" alt=\"Discord\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fbrowser-act\u002Fskills\u002Fstargazers\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbrowser-act\u002Fskills?style=social\" alt=\"GitHub Stars\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fbrowser-act\u002Fskills\u002Fblob\u002Fmain\u002FLICENSE\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-blue.svg\" alt=\"License\">\u003C\u002Fa>\n    \u003Cbr>\u003Cbr>\n    \u003Ca href=\"https:\u002F\u002Fwww.browseract.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FWebsite-BrowserAct.com-success\" alt=\"Website\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fx.com\u002Fbrowseract\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FX-browseract-000000?style=flat&logo=x&logoColor=white\" alt=\"X (Twitter)\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fwww.linkedin.com\u002Fcompany\u002Fbrowseract\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLinkedIn-BrowserAct-0A66C2?style=flat&logo=linkedin&logoColor=white\" alt=\"LinkedIn\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fwww.youtube.com\u002F@browseract\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FYouTube-@browseract-FF0000?style=flat&logo=youtube&logoColor=white\" alt=\"YouTube\">\u003C\u002Fa>\n  \u003C\u002Fp>\n\u003C\u002Fdiv>\n\n---\n\n# 🔥 Help your AI agent finish more browser tasks.\n\nA collection of Browser Skills for OPENCLAW-style agents and Skill workflows — built for runs that break on verification, redirects, unstable sessions, or actions that stop working mid-flow.\n\n---\n\n## ❓ What actually happens when most agents hit a real website\n\n* The task starts fine — then Cloudflare appears and everything stops\n* The page loads, but clicks stop working halfway through\n* Login works, then the session quietly dies on the next redirect\n* The agent runs for minutes, then returns nothing. Sound familiar?\n\n**If any of these have killed your workflow, BrowserAct is for you.**\n\n---\n\n## 🎯 Two Ways to Give Your Agent Browser Superpowers\n\nBrowserAct ships two flagship products. Pick the one that fits your problem — or use them together.\n\n### 💻 `browser-act` CLI — Real-time browser control\n\nDrives a real, stealth-capable Chrome from the command line. Your agent navigates, clicks, fills forms, handles captchas, and operates websites just like a human would.\n\n```bash\nnpx skills add browser-act\u002Fskills --skill browser-act\n```\n\n### 🔨 Skill Forge — Turn any website into a reusable Skill\n\nA meta-Skill that explores a site once, then generates a deploy-ready Skill package (SKILL.md + Python scripts). Every subsequent call skips the exploration cost.\n\n```bash\nnpx skills add browser-act\u002Fskills --skill browser-act-skill-forge\n```\n\n### Which one do I need?\n\n| Use `browser-act` CLI when... | Use Skill Forge when... |\n| :--- | :--- |\n| You want the agent to operate a website live | You need to extract or act repeatedly at scale |\n| One-off tasks, ad-hoc browsing | Batch jobs, recurring workflows, zero-maintenance automation |\n| Login flows, captchas, multi-account demos | Hundreds\u002Fthousands of records, cross-pagination, recurring monitoring |\n\n---\n\n## 💻 `browser-act` CLI\n\nFast, persistent browser automation from the command line:\n\n```bash\nbrowser-act navigate https:\u002F\u002Fexample.com  # Navigate to URL\nbrowser-act state                         # See clickable elements\nbrowser-act click 5                       # Click element by index\nbrowser-act input 3 \"Hello\"               # Click and type text\nbrowser-act screenshot page.png           # Take screenshot\n```\n\nIt supports two browser types (Stealth and Real Chrome) and provides scriptable commands for navigation, page interaction, data extraction, and session management.\n\n### ⚡ What makes it different\n\nMost AI agents get blocked by Cloudflare, CAPTCHAs, or login walls. `browser-act` provides the missing infrastructure to let your agent browse like a real human.\n\n| | |\n| :--- | :--- |\n| 🛡️ **Anti-Detection Stealth** | Bypasses Cloudflare, reCAPTCHA, Datadome, and more. Authentic browser fingerprints — your Agent won't get blocked. |\n| 🔗 **Real Chrome Control** | Drive your existing Chrome with all logins, cookies, and extensions. No re-authentication needed. |\n| 🚀 **Parallel Execution** | Run multiple stealth browsers concurrently — each with independent fingerprints, proxies, and sessions. |\n| 🤖 **Captcha Solving** | Built-in automatic captcha solving. No third-party services or manual intervention required. |\n| 📉 **Low Token Noise** | Strips 90% of junk HTML before feeding to the LLM. Save money, get faster answers. |\n| 🔒 **Proxy & Privacy Modes** | Per-browser proxy support (HTTP\u002FSOCKS5) and privacy mode for fresh environments on every launch. |\n\n*These aren't features for edge cases. They're fixes for the exact failures that stop most agents cold.*\n\n---\n\n## 🔨 Skill Forge\n\nEvery time you ask an AI agent to batch-extract from a new website, it starts from scratch — different path each time, different failure modes, unreliable at scale. If the site changes layout, the whole thing breaks.\n\nSkill Forge fixes that: **explore once, reuse forever.**\n\n**How it works:**\n\n1. **Describe** — tell the agent the target site and what data to extract or what action to perform\n2. **Explore** — Skill Forge automatically discovers API endpoints (preferred) or DOM patterns\n3. **Generate** — produces deploy-ready Skill files with business parameters extracted as CLI arguments\n4. **Auto-test** — a sub-agent verifies the Skill end-to-end and self-fixes failures until passing\n5. **Deploy & Scale** — run 500 or 5,000 records through the same stable path, anytime\n\n[**View Skill Forge →**](https:\u002F\u002Fgithub.com\u002Fbrowser-act\u002Fskills\u002Ftree\u002Fmain\u002Fbrowser-act-skill-forge)\n\n---\n\n## 🚀 Quick Start\n\n### Option A: Live browser control with `browser-act` CLI\n\n```bash\nnpx skills add browser-act\u002Fskills --skill browser-act\n```\n\nThen tell your agent:\n> *\"Go to Amazon.com, find the top 10 best-selling products in the mouse category, and save them to a markdown file.\"*\n\n### Option B: Forge a reusable Skill with Skill Forge\n\n```bash\nnpx skills add browser-act\u002Fskills --skill browser-act-skill-forge\n```\n\nThen tell your agent:\n> *\"Forge a Skill that extracts job listings from LinkedIn — title, company, salary, URL. I'll run it on 300 keywords later.\"*\n\nYour agent will prompt you to register and get a **free API Key** interactively if it encounters advanced anti-bot protections.\n\n---\n\n## ✨ Cross-Platform Compatibility\n\n**🚀 Works Seamlessly Across All Major AI Assistants**\n\nBrowserAct skills are designed to work **powerfully and reliably** on all leading AI coding platforms:\n\n| Platform | Status | Installation |\n| :--- | :--- | :--- |\n| **Claude Code** | ✅ Fully Supported | Native skill support |\n| **Cursor** | ✅ Fully Supported | Works out of the box |\n| **VS Code** | ✅ Fully Supported | Works out of the box |\n| **OpenCode** | ✅ Fully Supported | Direct integration |\n| **OpenClaw** | ✅ Fully Supported | Compatible |\n| **Codex** | ✅ Fully Supported | Compatible |\n| **Gemini CLI** | ✅ Fully Supported | Compatible |\n\n**Key Benefits:**\n- ✅ **Stable & Reliable:** No crashes, no unexpected behavior.\n- ✅ **Plug & Play:** Works immediately after installation.\n- ✅ **Cross-Platform:** Consistent performance across all AI assistants.\n\n---\n\n## 📦 Core Skills Catalog\n\nThis repository contains ready-to-use scenario skills covering E-commerce scraping, social media monitoring, lead generation, and more. \n\nHere are some of the highlighted skills you can use right away:\n\n- **Amazon ASIN Lookup Skill**: Extract structured product details from Amazon using ASIN.\n- **Amazon Best Selling Products Finder**: Extract best-selling product data based on keywords.\n- **Google News API Skill**: Track industry trends and breaking news.\n- **Google Maps API Skill**: Scrape local business data and contact info.\n- **YouTube Transcript Extractor**: Extract transcripts and metadata from YouTube videos automatically.\n\n[**Browse all available Skills in the `solutions` directory →**](https:\u002F\u002Fgithub.com\u002Fbrowser-act\u002Fskills\u002Ftree\u002Fmain\u002Fsolutions)\n\n*Works with any scenario. Can't find what you need? [Join Discord](https:\u002F\u002Fdiscord.com\u002Finvite\u002FUpnCKd7GaU) and request a Skill.*\n\n---\n\n\n## 💖 Support the Project\n\nBrowserAct Skills is **free and open source**. If it saves you time, please give us a ⭐ **Star** — it keeps the project alive and helps us ship more skills.\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fbrowser-act\u002Fskills\u002Fstargazers\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbrowser-act\u002Fskills?style=social\" alt=\"GitHub Stars\">\n\u003C\u002Fa>\n\n🎁 **Bonus:** Once you star the repository, you can join our [Discord](https:\u002F\u002Fdiscord.com\u002Finvite\u002FUpnCKd7GaU) and post in the `#claim-500-credits` channel to receive **500 free credits**!\n\n### 🤝 Community & Support\n- 💬 [Join our Discord](https:\u002F\u002Fdiscord.com\u002Finvite\u002FUpnCKd7GaU)\n- 📖 [Read the Docs](https:\u002F\u002Fdocs.browseract.com)\n- 🐛 [Report an Issue](https:\u002F\u002Fgithub.com\u002Fbrowser-act\u002Fskills\u002Fissues)\n- 🌐 [BrowserAct Website](https:\u002F\u002Fwww.browseract.com)\n\n\u003Cp align=\"center\">\u003Cem>Built with ❤️ by the BrowserAct Team\u003C\u002Fem>\u003C\u002Fp>\n","BrowserAct Skills 是一个让AI能够浏览网页、抓取并从复杂网站中提取结构化数据的工具，具有更快的执行速度、更低的成本和更可靠的结果。其核心功能包括通过命令行实时控制隐身能力的Chrome浏览器，以及将任意网站转换为可复用Skill的能力，从而降低后续操作成本。该项目使用Python编写，并支持自动化处理验证码、表单填写等任务。适用于需要频繁与动态网页交互或进行大规模数据抓取的场景，如市场调研、价格监控等领域。","2026-06-11 03:49:47","high_star"]