[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M72T0E807B2VYC69YAYDPX3E":3,"skill-same:01M72T0E807B2VYC69YAYDPX3E":48},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":27,"commitSha":28,"skillDirectory":29,"forksCount":14,"repositoryHtmlUrl":30,"repositoryHomepageUrl":30,"ownerHtmlUrl":31,"ownerBlogUrl":6,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M72T0E807B2VYC69YAYDPX3E","browser-automation",null,"Create AI agents that control web pages using natural language. Based on Alibaba's page-agent architecture. Use when building browser AI copilots, web automation agents, or DOM-based tool systems.","iberi22\u002Fskills","iberi22","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F10615454?v=4","01M72SQ2NXAA3GW0FSH053A5ES","Brahyan Belalcazar","01M72SQ2PYCH0SSA1K61NGEHDV",0,"MIT","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fgeneral\u002Fbrowser-automation\u002FSKILL.md",[],[],"2026-11-10 22:39:42",false,"---\nname: browser-automation\ndescription: Create AI agents that control web pages using natural language. Based on Alibaba's page-agent architecture. Use when building browser AI copilots, web automation agents, or DOM-based tool systems.\nversion: \"1.0.0\"\nupdated: \"2026-04-23\"\nlicense: MIT\nargument-hint: 'Describe the automation task or browser agent to build'\n---\n\n# Browser Automation Agent\n\nBuild AI agents that control web interfaces with natural language — no browser extensions, no Python, no headless browsers. Everything happens in-page.\n\n## Architecture Pattern (page-agent style)\n\n```\nUser Action (natural language)\n    ↓\nAgent (LLM + Tools)\n    ↓\nPageController (DOM operations)\n    ↓\nWeb Page (actual browser DOM)\n```\n\n## Core Components\n\n### 1. PageController\nHandles all DOM operations — no LLM dependency.\n\n```typescript\n\u002F\u002F Core methods every PageController must implement\nclass PageController {\n  async updateTree()           \u002F\u002F Extract live DOM\n  async getSimplifiedHTML()    \u002F\u002F Dehydrate for LLM\n  async getPageInfo()          \u002F\u002F Page metadata\n  async clickElement(index)    \u002F\u002F Indexed element interaction\n  async inputText(index, text) \u002F\u002F Form input\n  async scroll(direction)      \u002F\u002F Navigation\n}\n```\n\n### 2. Agent Brain (LLM Client)\nStateless LLM client with retry logic and reflection-before-action.\n\n```typescript\ninterface AgentConfig {\n  model: string\n  baseURL: string\n  apiKey: string\n  language: string\n}\n```\n\n### 3. Tool System\nTools are defined as functions that the LLM can call. Each tool:\n1. Calls PageController for DOM operations\n2. Returns structured result to LLM\n3. LLM decides next action\n\n## Key Files to Create\n\n| File | Purpose |\n|------|---------|\n| `PageController.ts` | DOM extraction + element operations |\n| `Agent.ts` | LLM client + tool executor + ReAct loop |\n| `tools\u002Findex.ts` | Tool definitions |\n| `prompts\u002Fsystem.md` | System prompt for the agent |\n| `types.ts` | TypeScript interfaces |\n\n## DOM Pipeline\n\n```\n1. Live DOM → FlatDomTree (page-controller)\n2. FlatDomTree → Simplified text (dehydration)\n3. Simplified text → LLM (action plan)\n4. Action plan → PageController operations (by index)\n```\n\n## Quick Start Pattern\n\n```typescript\nimport { PageAgent } from 'page-agent'\n\nconst agent = new PageAgent({\n  model: 'qwen3.5-plus',\n  baseURL: 'https:\u002F\u002Fdashscope.aliyuncs.com\u002Fcompatible-mode\u002Fv1',\n  apiKey: process.env.API_KEY,\n  language: 'en-US',\n})\n\nawait agent.execute('Click the login button')\n```\n\n## Monorepo Structure Pattern\n\n```\npackages\u002F\n├── core\u002F                    # Agent logic without UI (stateless)\n├── page-controller\u002F         # DOM operations (no LLM dependency)\n├── llms\u002F                    # LLM client with retry\u002Freflection\n├── ui\u002F                      # Optional panel\u002Foverlay\n└── page-agent\u002F              # Main entry (core + controller + ui)\n```\n\n## Best Practices\n\n1. **Indexed operations** — Don't pass DOM elements to LLM, use integer indices\n2. **Simplified HTML** — Remove noise (styles, scripts) before sending to LLM\n3. **Retry logic** — LLM calls fail; implement exponential backoff\n4. **Visual feedback** — Optional mask overlay shows agent actions to user\n5. **No screenshots** — Text-based DOM manipulation only (faster, cheaper)\n\n## When to Use\n\n- ✅ Building an AI copilot for a web app\n- ✅ Automating form filling workflows\n- ✅ Creating accessibility overlays\n- ✅ Multi-page browser agents\n- ❌ Server-side scraping (use puppeteer\u002Fplaywright instead)\n- ❌ Complex visual tasks requiring screenshots\n\n## Integration Options\n\n| Approach | Use Case |\n|----------|----------|\n| NPM package | Full control, custom UI |\n| Chrome extension | Multi-page tasks |\n| MCP Server | External agent control |\n| IIFE script tag | Quick integration |\n\n## Not the same as agentic testing\n\nThis skill builds agents that operate a page for a user (copilots, DOM tools). To TEST an app with an agent —\nnatural-language steps, replay cache, exploratory bug bashes with repro tests — use `agentic-e2e`\n(tester-army\u002Fe2e) instead.\n","# Browser Automation Agent\n\nBuild AI agents that control web interfaces with natural language — no browser extensions, no Python, no headless browsers. Everything happens in-page.\n\n## Architecture Pattern (page-agent style)\n\n```\nUser Action (natural language)\n    ↓\nAgent (LLM + Tools)\n    ↓\nPageController (DOM operations)\n    ↓\nWeb Page (actual browser DOM)\n```\n\n## Core Components\n\n### 1. PageController\nHandles all DOM operations — no LLM dependency.\n\n```typescript\n\u002F\u002F Core methods every PageController must implement\nclass PageController {\n  async updateTree()           \u002F\u002F Extract live DOM\n  async getSimplifiedHTML()    \u002F\u002F Dehydrate for LLM\n  async getPageInfo()          \u002F\u002F Page metadata\n  async clickElement(index)    \u002F\u002F Indexed element interaction\n  async inputText(index, text) \u002F\u002F Form input\n  async scroll(direction)      \u002F\u002F Navigation\n}\n```\n\n### 2. Agent Brain (LLM Client)\nStateless LLM client with retry logic and reflection-before-action.\n\n```typescript\ninterface AgentConfig {\n  model: string\n  baseURL: string\n  apiKey: string\n  language: string\n}\n```\n\n### 3. Tool System\nTools are defined as functions that the LLM can call. Each tool:\n1. Calls PageController for DOM operations\n2. Returns structured result to LLM\n3. LLM decides next action\n\n## Key Files to Create\n\n| File | Purpose |\n|------|---------|\n| `PageController.ts` | DOM extraction + element operations |\n| `Agent.ts` | LLM client + tool executor + ReAct loop |\n| `tools\u002Findex.ts` | Tool definitions |\n| `prompts\u002Fsystem.md` | System prompt for the agent |\n| `types.ts` | TypeScript interfaces |\n\n## DOM Pipeline\n\n```\n1. Live DOM → FlatDomTree (page-controller)\n2. FlatDomTree → Simplified text (dehydration)\n3. Simplified text → LLM (action plan)\n4. Action plan → PageController operations (by index)\n```\n\n## Quick Start Pattern\n\n```typescript\nimport { PageAgent } from 'page-agent'\n\nconst agent = new PageAgent({\n  model: 'qwen3.5-plus',\n  baseURL: 'https:\u002F\u002Fdashscope.aliyuncs.com\u002Fcompatible-mode\u002Fv1',\n  apiKey: process.env.API_KEY,\n  language: 'en-US',\n})\n\nawait agent.execute('Click the login button')\n```\n\n## Monorepo Structure Pattern\n\n```\npackages\u002F\n├── core\u002F                    # Agent logic without UI (stateless)\n├── page-controller\u002F         # DOM operations (no LLM dependency)\n├── llms\u002F                    # LLM client with retry\u002Freflection\n├── ui\u002F                      # Optional panel\u002Foverlay\n└── page-agent\u002F              # Main entry (core + controller + ui)\n```\n\n## Best Practices\n\n1. **Indexed operations** — Don't pass DOM elements to LLM, use integer indices\n2. **Simplified HTML** — Remove noise (styles, scripts) before sending to LLM\n3. **Retry logic** — LLM calls fail; implement exponential backoff\n4. **Visual feedback** — Optional mask overlay shows agent actions to user\n5. **No screenshots** — Text-based DOM manipulation only (faster, cheaper)\n\n## When to Use\n\n- ✅ Building an AI copilot for a web app\n- ✅ Automating form filling workflows\n- ✅ Creating accessibility overlays\n- ✅ Multi-page browser agents\n- ❌ Server-side scraping (use puppeteer\u002Fplaywright instead)\n- ❌ Complex visual tasks requiring screenshots\n\n## Integration Options\n\n| Approach | Use Case |\n|----------|----------|\n| NPM package | Full control, custom UI |\n| Chrome extension | Multi-page tasks |\n| MCP Server | External agent control |\n| IIFE script tag | Quick integration |\n\n## Not the same as agentic testing\n\nThis skill builds agents that operate a page for a user (copilots, DOM tools). To TEST an app with an agent —\nnatural-language steps, replay cache, exploratory bug bashes with repro tests — use `agentic-e2e`\n(tester-army\u002Fe2e) instead.\n",{"name":5,"license":15,"updated":24,"version":25,"description":7,"argument-hint":26},"2026-04-23","1.0.0","Describe the automation task or browser agent to build","main","688810f2886d212632acf81e532b24e483dd8b98","skills\u002Fgeneral\u002Fbrowser-automation","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills","https:\u002F\u002Fgithub.com\u002Fiberi22","Senior WEB Full-Stack","2026-10-12 02:46:21",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add iberi22\u002Fskills --skill browser-automation",{"tool":40,"command":41},"bunx","bunx skills add iberi22\u002Fskills --skill browser-automation",{"tool":43,"command":44},"pnpm","pnpm dlx skills add iberi22\u002Fskills --skill browser-automation","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fgeneral\u002Fbrowser-automation\u002FSKILL.md",true,[],[49,55,62,68,74,80],{"skillNo":50,"name":51,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":52,"categories":53,"keywords":54,"publishedAt":19},"01M72TWQEQD46ZC9TRBAABF78P","web-design-guidelines","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fweb\u002Fweb-design-guidelines\u002FSKILL.md",[],[],{"skillNo":56,"name":57,"nameZh":6,"description":58,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":59,"categories":60,"keywords":61,"publishedAt":19},"01M72TWMQFTYD08DGXVAXRPPEB","blocked-page-recovery","Use when a fetch fails: 403\u002F429, paywall, WAF, bot wall. Recovers the page via archives and provenance-checked fallbacks.","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Fweb\u002Fblocked-page-recovery\u002FSKILL.md",[],[],{"skillNo":63,"name":64,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":65,"categories":66,"keywords":67,"publishedAt":19},"01M72TVQMS3HCTX18C8T97WTQQ","studio-coach","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftools\u002Fstudio-coach\u002FSKILL.md",[],[],{"skillNo":69,"name":70,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":19},"01M72TVNA2PR9RZW0ZAXYP1G6J","src-generator","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftools\u002Fsrc-generator\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":19},"01M72TVEJHHB6SAT3R3TY9MJKS","sqlite-pro","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftools\u002Fsqlite-pro\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":19},"01M72TVC2EVBHQHCX96VTTYKB8","rigor-evaluator","Use when you need a quick heuristic rigor score (10 criteria, 0-2 each) for the features in a .gitcore\u002Ffeatures.json file, to decide which features are ready to ship.","https:\u002F\u002Fgithub.com\u002Fiberi22\u002Fskills\u002Fblob\u002Fmain\u002Fskills\u002Ftools\u002Frigor-evaluator\u002FSKILL.md",[],[]]