[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AE10WVZKMHC9V9QRRWHS2C":3,"skill-same:01M7AE10WVZKMHC9V9QRRWHS2C":44},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M7AE10WVZKMHC9V9QRRWHS2C","nodetool-browser-agent",null,"Create browser automation agents that navigate websites, extract data, fill forms, and perform multi-step web tasks using natural language instructions. Use when user asks to automate browsing, scrape websites with AI, build a web agent, or perform complex browser interactions.","hanchenhao\u002Fai-canvas","hanchenhao","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F15713703?v=4","01M7ADWCF5AE30N3MG7MDF3RV2","01M7ADWCFVJS4V8VD9KWJG22WR",1,"AGPL-3.0","https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fnodetool-browser-agent\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: nodetool-browser-agent\ndescription: Create browser automation agents that navigate websites, extract data, fill forms, and perform multi-step web tasks using natural language instructions. Use when user asks to automate browsing, scrape websites with AI, build a web agent, or perform complex browser interactions.\n---\n\nYou help users create NodeTool agents configured for browser automation — AI-powered web agents that navigate, interact with, and extract data from websites using natural-language task descriptions.\n\n# Architecture\n\nBrowser automation in NodeTool uses the **agent system**. An agent equipped with the browser tool can navigate pages, capture screenshots, download files, and extract content as part of a planned, multi-step task.\n\n```\nTask description → Agent → (Plan steps → Use browser tools → Extract\u002Finteract → Report) → Result\n```\n\n# Available Browser Tools\n\nThese are the web-related tools the agent CLI resolves from the YAML `tools:` list:\n\n| Tool name | Description |\n|-----------|-------------|\n| `browser` | Fetch and render web page content (executes JavaScript), extract text\u002Flinks |\n| `screenshot` | Capture a screenshot of a page |\n| `download_file` | Download a file from a URL into the workspace |\n| `google_search` | Web search to discover URLs |\n| `google_news` \u002F `google_images` | News \u002F image search |\n| `http_request` | Raw HTTP GET\u002FPOST for APIs and simple fetches |\n| `openai_web_search` | OpenAI-hosted web search (needs an OpenAI key) |\n\n> There are no `dom_examine`, `dom_search`, or `dom_extract` tools. The `browser`\n> tool handles rendering and content extraction; instruct the agent in natural\n> language (e.g. \"extract every product title and price\") and pair it with\n> `write_file` to save structured results.\n\n# Agent YAML Config for Browser Automation\n\n```yaml\nname: browser-agent\ndescription: AI-powered browser automation agent\n\nsystem_prompt: |\n  You are a web automation agent. You navigate websites, interact with pages,\n  and extract information based on user instructions.\n\n  Workflow:\n  1. Analyze the task to determine what pages to visit\n  2. Use the browser tool to fetch and read pages\n  3. Extract the relevant content\n  4. Report findings in a structured format\n\n  Guidelines:\n  - Start by browsing the target URL to understand page structure\n  - Use screenshot to verify visual state when needed\n  - Validate extracted data before reporting\n\nmodel:\n  provider: openai\n  id: gpt-5.4\n\nplanning_agent:\n  enabled: true\n  model:\n    provider: openai\n    id: gpt-5.4-mini\n\ntools:\n  - browser\n  - screenshot\n  - download_file\n  - http_request\n  - google_search\n  - write_file\n\nmax_steps: 15\n```\n\n# Quick Recipes\n\n## Web Scraper Agent\n```yaml\nname: web-scraper\ndescription: Extract structured data from websites\nsystem_prompt: |\n  You extract structured data from websites. For each URL:\n  1. Browse the page to understand its structure\n  2. Identify the data elements (titles, prices, dates)\n  3. Extract them into structured JSON\n  4. Save results with write_file\nmodel: { provider: openai, id: gpt-5.4 }\ntools: [browser, write_file]\nmax_steps: 20\n```\n\n## Research Agent with Browser\n```yaml\nname: web-researcher\ndescription: Research topics using web search and browsing\nsystem_prompt: |\n  You are a research agent. For each topic:\n  1. Search the web for relevant sources\n  2. Browse promising results for detailed content\n  3. Extract key information and verify facts\n  4. Compile findings into a structured report\nmodel: { provider: openai, id: gpt-5.4 }\ntools: [google_search, browser, write_file]\nmax_steps: 15\n```\n\n## Price Comparison Agent\n```yaml\nname: price-compare\ndescription: Compare prices across websites\nsystem_prompt: |\n  You compare product prices across websites:\n  1. Search for the product on each site\n  2. Extract price, availability, and shipping info\n  3. Create a comparison table\n  4. Recommend the best deal\nmodel: { provider: openai, id: gpt-5.4 }\ntools: [google_search, browser, write_file]\nmax_steps: 20\n```\n\n# CLI Usage\n\n```bash\n# Run a browser task (objective via flag, stdin, or `objective:` in the YAML)\nnodetool agent run browser-agent.yaml \\\n  --objective \"Go to example.com and extract all product names and prices\"\n\n# Research task — final answer to stdout, trace to stderr\nnodetool agent run web-researcher.yaml \\\n  --objective \"Research the latest developments in WebAssembly\" \\\n  > research-report.md\n```\n\n# Browser Agent as a Workflow Node\n\nFor visual workflows (and the DSL), there is a dedicated `BrowserAgent` node\n(`agents.browserAgent` in `@nodetool-ai\u002Fdsl`). It runs a browsing agent inside a\ngraph and returns the extracted text. Use this when browsing is one step of a\nlarger pipeline rather than a standalone CLI run.\n\n# Programmatic Usage (TypeScript)\n\n```typescript\nimport { Agent } from \"@nodetool-ai\u002Fagents\";\nimport { BrowserTool, ScreenshotTool, DownloadFileTool } from \"@nodetool-ai\u002Fagents\";\nimport { ProcessingContext, FileStorageAdapter } from \"@nodetool-ai\u002Fruntime\";\n\nconst agent = new Agent({\n  name: \"browser-agent\",\n  objective: \"Extract product listings from example.com\",\n  provider: openaiProvider,\n  model: \"gpt-5.4\",\n  tools: [new BrowserTool(), new ScreenshotTool(), new DownloadFileTool()],\n  workspace: \"\u002Ftmp\u002Fbrowser-output\",\n  maxSteps: 15,\n});\n\nconst ctx = new ProcessingContext({\n  jobId: `browser-${Date.now()}`,\n  userId: \"1\",\n  workspaceDir: \"\u002Ftmp\u002Fbrowser-output\",\n  workspaceStorage: new FileStorageAdapter(\"\u002Ftmp\u002Fbrowser-output\"),\n});\n\nfor await (const message of agent.execute(ctx)) {\n  if (message.type === \"chunk\") process.stdout.write(message.content);\n}\n```\n\n# Tips\n\n- **Describe extraction in natural language** — the `browser` tool returns page content; let the model parse it. Pair with `write_file` to persist results.\n- **Use `screenshot`** to debug visual state or verify a page loaded.\n- **Set `max_steps` higher** (15-20) for multi-page tasks.\n- **Combine with `google_search`** when the agent needs to discover URLs first.\n- **Use `http_request`** for JSON APIs — it's faster and cheaper than full page rendering.\n","You help users create NodeTool agents configured for browser automation — AI-powered web agents that navigate, interact with, and extract data from websites using natural-language task descriptions.\n\n# Architecture\n\nBrowser automation in NodeTool uses the **agent system**. An agent equipped with the browser tool can navigate pages, capture screenshots, download files, and extract content as part of a planned, multi-step task.\n\n```\nTask description → Agent → (Plan steps → Use browser tools → Extract\u002Finteract → Report) → Result\n```\n\n# Available Browser Tools\n\nThese are the web-related tools the agent CLI resolves from the YAML `tools:` list:\n\n| Tool name | Description |\n|-----------|-------------|\n| `browser` | Fetch and render web page content (executes JavaScript), extract text\u002Flinks |\n| `screenshot` | Capture a screenshot of a page |\n| `download_file` | Download a file from a URL into the workspace |\n| `google_search` | Web search to discover URLs |\n| `google_news` \u002F `google_images` | News \u002F image search |\n| `http_request` | Raw HTTP GET\u002FPOST for APIs and simple fetches |\n| `openai_web_search` | OpenAI-hosted web search (needs an OpenAI key) |\n\n> There are no `dom_examine`, `dom_search`, or `dom_extract` tools. The `browser`\n> tool handles rendering and content extraction; instruct the agent in natural\n> language (e.g. \"extract every product title and price\") and pair it with\n> `write_file` to save structured results.\n\n# Agent YAML Config for Browser Automation\n\n```yaml\nname: browser-agent\ndescription: AI-powered browser automation agent\n\nsystem_prompt: |\n  You are a web automation agent. You navigate websites, interact with pages,\n  and extract information based on user instructions.\n\n  Workflow:\n  1. Analyze the task to determine what pages to visit\n  2. Use the browser tool to fetch and read pages\n  3. Extract the relevant content\n  4. Report findings in a structured format\n\n  Guidelines:\n  - Start by browsing the target URL to understand page structure\n  - Use screenshot to verify visual state when needed\n  - Validate extracted data before reporting\n\nmodel:\n  provider: openai\n  id: gpt-5.4\n\nplanning_agent:\n  enabled: true\n  model:\n    provider: openai\n    id: gpt-5.4-mini\n\ntools:\n  - browser\n  - screenshot\n  - download_file\n  - http_request\n  - google_search\n  - write_file\n\nmax_steps: 15\n```\n\n# Quick Recipes\n\n## Web Scraper Agent\n```yaml\nname: web-scraper\ndescription: Extract structured data from websites\nsystem_prompt: |\n  You extract structured data from websites. For each URL:\n  1. Browse the page to understand its structure\n  2. Identify the data elements (titles, prices, dates)\n  3. Extract them into structured JSON\n  4. Save results with write_file\nmodel: { provider: openai, id: gpt-5.4 }\ntools: [browser, write_file]\nmax_steps: 20\n```\n\n## Research Agent with Browser\n```yaml\nname: web-researcher\ndescription: Research topics using web search and browsing\nsystem_prompt: |\n  You are a research agent. For each topic:\n  1. Search the web for relevant sources\n  2. Browse promising results for detailed content\n  3. Extract key information and verify facts\n  4. Compile findings into a structured report\nmodel: { provider: openai, id: gpt-5.4 }\ntools: [google_search, browser, write_file]\nmax_steps: 15\n```\n\n## Price Comparison Agent\n```yaml\nname: price-compare\ndescription: Compare prices across websites\nsystem_prompt: |\n  You compare product prices across websites:\n  1. Search for the product on each site\n  2. Extract price, availability, and shipping info\n  3. Create a comparison table\n  4. Recommend the best deal\nmodel: { provider: openai, id: gpt-5.4 }\ntools: [google_search, browser, write_file]\nmax_steps: 20\n```\n\n# CLI Usage\n\n```bash\n# Run a browser task (objective via flag, stdin, or `objective:` in the YAML)\nnodetool agent run browser-agent.yaml \\\n  --objective \"Go to example.com and extract all product names and prices\"\n\n# Research task — final answer to stdout, trace to stderr\nnodetool agent run web-researcher.yaml \\\n  --objective \"Research the latest developments in WebAssembly\" \\\n  > research-report.md\n```\n\n# Browser Agent as a Workflow Node\n\nFor visual workflows (and the DSL), there is a dedicated `BrowserAgent` node\n(`agents.browserAgent` in `@nodetool-ai\u002Fdsl`). It runs a browsing agent inside a\ngraph and returns the extracted text. Use this when browsing is one step of a\nlarger pipeline rather than a standalone CLI run.\n\n# Programmatic Usage (TypeScript)\n\n```typescript\nimport { Agent } from \"@nodetool-ai\u002Fagents\";\nimport { BrowserTool, ScreenshotTool, DownloadFileTool } from \"@nodetool-ai\u002Fagents\";\nimport { ProcessingContext, FileStorageAdapter } from \"@nodetool-ai\u002Fruntime\";\n\nconst agent = new Agent({\n  name: \"browser-agent\",\n  objective: \"Extract product listings from example.com\",\n  provider: openaiProvider,\n  model: \"gpt-5.4\",\n  tools: [new BrowserTool(), new ScreenshotTool(), new DownloadFileTool()],\n  workspace: \"\u002Ftmp\u002Fbrowser-output\",\n  maxSteps: 15,\n});\n\nconst ctx = new ProcessingContext({\n  jobId: `browser-${Date.now()}`,\n  userId: \"1\",\n  workspaceDir: \"\u002Ftmp\u002Fbrowser-output\",\n  workspaceStorage: new FileStorageAdapter(\"\u002Ftmp\u002Fbrowser-output\"),\n});\n\nfor await (const message of agent.execute(ctx)) {\n  if (message.type === \"chunk\") process.stdout.write(message.content);\n}\n```\n\n# Tips\n\n- **Describe extraction in natural language** — the `browser` tool returns page content; let the model parse it. Pair with `write_file` to persist results.\n- **Use `screenshot`** to debug visual state or verify a page loaded.\n- **Set `max_steps` higher** (15-20) for multi-page tasks.\n- **Combine with `google_search`** when the agent needs to discover URLs first.\n- **Use `http_request`** for JSON APIs — it's faster and cheaper than full page rendering.\n",{"name":5,"description":7},"main","4ef72a828b24c99c49109033c7ec504c34f3bd17",".claude\u002Fskills\u002Fnodetool-browser-agent",0,"https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas","https:\u002F\u002Fgithub.com\u002Fhanchenhao","2026-08-18 16:41:12",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add hanchenhao\u002Fai-canvas --skill nodetool-browser-agent",{"tool":36,"command":37},"bunx","bunx skills add hanchenhao\u002Fai-canvas --skill nodetool-browser-agent",{"tool":39,"command":40},"pnpm","pnpm dlx skills add hanchenhao\u002Fai-canvas --skill nodetool-browser-agent","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas\u002Fblob\u002Fmain\u002F.claude\u002Fskills\u002Fnodetool-browser-agent\u002FSKILL.md",true,[],[45,52,59,66,73,80],{"skillNo":46,"name":47,"nameZh":6,"description":48,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":49,"categories":50,"keywords":51,"publishedAt":18},"01M7AE5NP5AWGV47WS3F37VPZ1","sandbox-zip","Read and write zip archives in a Code node or CodeAct action, with fflate running on the host under a 50 MB inflation cap","https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas\u002Fblob\u002Fmain\u002Fpackages\u002Fsandbox-packs\u002Fsandbox-zip\u002FSKILL.md",[],[],{"skillNo":53,"name":54,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":56,"categories":57,"keywords":58,"publishedAt":18},"01M7AE5JP0CG2S114Q246RC6MH","sandbox-yaml","Read and write YAML inside a Code node or CodeAct action, with js-yaml running in the guest","https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas\u002Fblob\u002Fmain\u002Fpackages\u002Fsandbox-packs\u002Fsandbox-yaml\u002FSKILL.md",[],[],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":18},"01M7AE5GJ9AV36WM1DPW1878VM","sandbox-xml","Parse XML feeds and sitemaps in a Code node or CodeAct action, with fast-xml-parser running on the host","https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas\u002Fblob\u002Fmain\u002Fpackages\u002Fsandbox-packs\u002Fsandbox-xml\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":18},"01M7AE5DMYDAGJW27G8KKTZ2BN","sandbox-xlsx","Read and write Excel workbooks in a Code node or CodeAct action, with exceljs running on the host","https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas\u002Fblob\u002Fmain\u002Fpackages\u002Fsandbox-packs\u002Fsandbox-xlsx\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":18},"01M7AE5BARJSFCT296SEMYDSJQ","sandbox-twilio","Send SMS and WhatsApp messages and look up numbers from a Code node or CodeAct action, with NodeTool's Twilio request builder running on the host","https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas\u002Fblob\u002Fmain\u002Fpackages\u002Fsandbox-packs\u002Fsandbox-twilio\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":18},"01M7AE57C5APRDF96NB3G0WQMW","sandbox-tfjs","Classify and embed images, detect objects, and answer from a passage in a Code node or CodeAct action, with TensorFlow.js running on the host","https:\u002F\u002Fgithub.com\u002Fhanchenhao\u002Fai-canvas\u002Fblob\u002Fmain\u002Fpackages\u002Fsandbox-packs\u002Fsandbox-tfjs\u002FSKILL.md",[],[]]