使用GPT-4o问答助手分析文档和网页内容
中级
这是一个自动化工作流，包含 12 个节点。主要使用 If, Code, HttpRequest, ReadBinaryFile, Agent 等节点。 使用GPT-4o问答助手分析文档和网页内容
前置要求
- •可能需要目标 API 的认证凭证
- •OpenAI API Key
分类
-
工作流预览
可视化展示节点连接关系，支持缩放和平移
导出工作流
复制以下 JSON 配置到 n8n 导入，即可使用此工作流
{
"id": "NMAA4tOidWSLW3On",
"meta": {
"instanceId": "4a409a91d2a3d01c353bbc5fe897fb8ad93edeeea7f4f1264dcf7a9735497f14",
"templateCredsSetupCompleted": true
},
"name": "文档分析器和问答",
"tags": [],
"nodes": [
{
"id": "8ec9d857-a965-47e0-a367-3172a1056232",
"name": "文档问答聊天",
"type": "@n8n/n8n-nodes-langchain.chatTrigger",
"position": [
192,
208
],
"webhookId": "simple-doc-analyzer-chat",
"parameters": {
"options": {}
},
"typeVersion": 1.1
},
{
"id": "aaaefabe-e8fe-45b0-b841-109afc0049b3",
"name": "解析文档和问题",
"type": "n8n-nodes-base.code",
"position": [
432,
208
],
"parameters": {
"jsCode": "const chatInput = $input.first().json.chatInput;\nconst parts = chatInput.split('|').map(part => part.trim());\n\nif (parts.length < 2) {\n throw new Error('Please provide input in format: \"document_path_or_url | your_question\"\\nExamples:\\n- \"/Users/docs/readme.md | What is this project about?\"\\n- \"https://docs.example.com/api | What are the endpoints?\"');\n}\n\nconst documentPath = parts[0];\nconst userQuestion = parts.slice(1).join('|').trim();\n\nif (!documentPath || documentPath.length < 3) {\n throw new Error('Please provide a valid document path or URL');\n}\n\nif (!userQuestion || userQuestion.length < 5) {\n throw new Error('Please provide a detailed question (minimum 5 characters)');\n}\n\nconst isUrl = documentPath.startsWith('http://') || documentPath.startsWith('https://');\n\nlet fileType = 'unknown';\nlet inputType = 'file';\n\nif (isUrl) {\n inputType = 'url';\n fileType = 'html';\n \n const urlParts = documentPath.split('.');\n if (urlParts.length > 1) {\n const possibleExtension = urlParts[urlParts.length - 1].split(/[?#]/)[0].toLowerCase();\n const supportedTypes = ['pdf', 'md', 'txt', 'doc', 'docx', 'json', 'yaml', 'yml'];\n if (supportedTypes.includes(possibleExtension)) {\n fileType = possibleExtension;\n }\n }\n} else {\n const fileExtension = documentPath.split('.').pop().toLowerCase();\n const supportedTypes = ['pdf', 'md', 'txt', 'doc', 'docx', 'json', 'yaml', 'yml'];\n \n if (!supportedTypes.includes(fileExtension)) {\n throw new Error(`Unsupported file type: ${fileExtension}. Supported types: ${supportedTypes.join(', ')}`);\n }\n \n fileType = fileExtension;\n}\n\nreturn {\n documentPath: documentPath,\n userQuestion: userQuestion,\n fileType: fileType,\n inputType: inputType,\n isUrl: isUrl,\n timestamp: new Date().toISOString()\n};"
},
"typeVersion": 2
},
{
"id": "8ad92e8a-54e5-4896-84aa-a054abc5c7f3",
"name": "文件路径检查",
"type": "n8n-nodes-base.if",
"position": [
672,
96
],
"parameters": {
"options": {},
"conditions": {
"options": {
"version": 1,
"leftValue": "",
"caseSensitive": true,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "file-condition",
"operator": {
"type": "boolean",
"operation": "equals"
},
"leftValue": "={{ $json.isUrl }}",
"rightValue": false
}
]
}
},
"typeVersion": 2
},
{
"id": "afd1bad4-bc51-4f3d-8099-98dd9de1aac4",
"name": "读取文档文件",
"type": "n8n-nodes-base.readBinaryFile",
"position": [
832,
96
],
"parameters": {
"filePath": "={{ $json.documentPath }}"
},
"typeVersion": 1
},
{
"id": "c7024fab-3b1c-4892-abab-152a91c94b53",
"name": "URL 检查",
"type": "n8n-nodes-base.if",
"position": [
672,
320
],
"parameters": {
"options": {},
"conditions": {
"options": {
"version": 1,
"leftValue": "",
"caseSensitive": true,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "url-condition",
"operator": {
"type": "boolean",
"operation": "equals"
},
"leftValue": "={{ $json.isUrl }}",
"rightValue": true
}
]
}
},
"typeVersion": 2
},
{
"id": "bdb7e2fd-1441-4cac-a66e-dc20d8f04aeb",
"name": "获取网页内容",
"type": "n8n-nodes-base.httpRequest",
"position": [
832,
320
],
"parameters": {
"url": "={{ $json.documentPath }}",
"options": {
"timeout": 30000,
"redirect": {
"redirect": {
"maxRedirects": 5
}
},
"response": {
"response": {
"responseFormat": "text"
}
}
},
"sendHeaders": true,
"headerParameters": {
"parameters": [
{
"name": "User-Agent",
"value": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
},
{
"name": "Accept",
"value": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8"
},
{
"name": "Accept-Language",
"value": "en-US,en;q=0.9"
}
]
}
},
"typeVersion": 4.2
},
{
"id": "9a96077e-2158-42c0-853f-3dd43188c28d",
"name": "提取文档内容",
"type": "n8n-nodes-base.extractFromFile",
"position": [
992,
96
],
"parameters": {
"options": {},
"operation": "text"
},
"typeVersion": 1
},
{
"id": "053e4752-6062-4641-8119-3d7aacbf6af1",
"name": "处理文档内容",
"type": "n8n-nodes-base.code",
"position": [
1200,
208
],
"parameters": {
"jsCode": "const fileType = $node['Parse Document & Question'].json.fileType;\nconst inputType = $node['Parse Document & Question'].json.inputType;\nconst isUrl = $node['Parse Document & Question'].json.isUrl;\nconst documentPath = $node['Parse Document & Question'].json.documentPath;\n\nlet extractedContent;\nlet processedContent;\n\nconst inputData = $input.first();\n\nif (isUrl) {\n extractedContent = inputData.json.body || inputData.json.data || inputData.json || '';\n if (fileType === 'html') {\n const htmlString = typeof extractedContent === 'object' ? JSON.stringify(extractedContent) : String(extractedContent || '');\n processedContent = htmlString\n .replace(/<script[^>]*>[\\s\\S]*?<\\/script>/gi, '')\n .replace(/<style[^>]*>[\\s\\S]*?<\\/style>/gi, '')\n .replace(/<[^>]*>/g, ' ')\n .replace(/ /g, ' ')\n .replace(/&/g, '&')\n .replace(/</g, '<')\n .replace(/>/g, '>')\n .replace(/"/g, '\"')\n .replace(/'/g, \"'\")\n .replace(/\\s+/g, ' ')\n .trim();\n } else {\n processedContent = typeof extractedContent === 'object' ? JSON.stringify(extractedContent) : String(extractedContent || '');\n }\n} else {\n extractedContent = inputData.json.data || inputData.json.text || inputData.json || inputData.binary?.data?.toString() || '';\n processedContent = typeof extractedContent === 'object' ? JSON.stringify(extractedContent) : String(extractedContent || '');\n}\nswitch(fileType) {\n case 'json':\n try {\n const jsonData = JSON.parse(processedContent);\n processedContent = JSON.stringify(jsonData, null, 2);\n } catch (e) {\n processedContent = processedContent;\n }\n break;\n \n case 'yaml':\n case 'yml':\n break;\n \n case 'md':\n processedContent = processedContent\n .replace(/^\\s*```[^\\n]*\\n/gm, '\\n--- CODE BLOCK ---\\n')\n .replace(/^\\s*```\\s*$/gm, '\\n--- END CODE BLOCK ---\\n');\n break;\n \n case 'html':\n break;\n \n default:\n processedContent = processedContent\n .replace(/\\s+/g, ' ')\n .trim();\n}\n\nif (!processedContent || processedContent.length < 10) {\n throw new Error(`Failed to extract meaningful content from ${isUrl ? 'web page' : 'document'}. Content may be empty, corrupted, or unsupported.`);\n}\n\nconst maxLength = 15000;\nconst originalLength = processedContent.length;\nif (processedContent.length > maxLength) {\n processedContent = processedContent.substring(0, maxLength) + '\\n\\n[Content truncated for processing...]';\n}\n\nreturn {\n documentContent: processedContent,\n documentPath: documentPath,\n fileType: fileType,\n inputType: inputType,\n isUrl: isUrl,\n contentLength: processedContent.length,\n isContentTruncated: originalLength > maxLength\n};"
},
"typeVersion": 2
},
{
"id": "ce10391d-1e7b-4a68-9708-690961633dd1",
"name": "OpenAI 文档分析器",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
1408,
368
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "gpt-4o"
},
"options": {}
},
"credentials": {
"openAiApi": {
"id": "cjWgmpi35pveOOWw",
"name": "OpenAi account"
}
},
"typeVersion": 1.2
},
{
"id": "62a54e68-13ae-468c-acf0-b164138a19c0",
"name": "分析文档并回答",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
1408,
208
],
"parameters": {
"text": "=You are an expert document analysis assistant. Your task is to carefully read and understand the provided content, then answer the user's question accurately and comprehensively with well-formatted, human-readable output.\n\n**Content Information:**\n- Source: {{ $node['Parse Document & Question'].json.isUrl ? 'Web URL' : 'Local File' }}\n- Path/URL: {{ $node['Parse Document & Question'].json.documentPath }}\n- Content Type: {{ $node['Process Document Content'].json.fileType }}\n- Input Type: {{ $node['Process Document Content'].json.inputType }}\n- Content Length: {{ $node['Process Document Content'].json.contentLength }} characters\n- Content Truncated: {{ $node['Process Document Content'].json.isContentTruncated }}\n\n**Content:**\n{{ $node['Process Document Content'].json.documentContent }}\n\n**User Question:**\n{{ $node['Parse Document & Question'].json.userQuestion }}\n\n**Instructions:**\n1. Start your response with a header showing the source and question\n2. **Carefully read and analyze** the entire content provided above\n3. **Understand the context** and structure of the content\n4. **Answer the user's question** based on the content\n5. **Format your response** for maximum readability using proper formatting\n6. **Use tables, bullet points, and structured layouts** when appropriate\n7. **If the answer is not in the content**, clearly state that the information is not available\n8. **Provide context** around your answer when helpful\n9. **For web content**, focus on the main content and ignore navigation/footer elements\n\n**Response Formatting Requirements:**\n- Use bullet points (•) for lists and key points\n- Create tables when presenting structured data or comparisons\n- Use headings and subheadings for organization\n- Include numbered steps for processes or procedures\n- Quote relevant sections with proper formatting\n- Use emojis sparingly for visual appeal and clarity\n\n**Response Structure:**\n\n# 📊 Document Analysis Report\n\n## 🎯 Query Information\n• **Question Asked:** {{ $node['Parse Document & Question'].json.userQuestion }}\n• **Source Type:** {{ $node['Parse Document & Question'].json.isUrl ? '🌐 Web Page' : '📄 Local File' }}\n• **Source:** {{ $node['Parse Document & Question'].json.documentPath }}\n• **Analysis Status:** ✅ Complete\n\n---\n\n## 📋 **Direct Answer**\n[Provide a clear, concise answer to the user's question]\n\n## 📖 **Key Information**\n• [Main point 1]\n• [Main point 2] \n• [Main point 3]\n\n## 📊 **Detailed Analysis**\n[Use tables, bullet points, or structured format as appropriate]\n\n| Aspect | Details |\n|--------|----------|\n| [Key 1] | [Value 1] |\n| [Key 2] | [Value 2] |\n\n## 🔍 **Supporting Evidence**\n> \"[Relevant quote from content]\"\n\n• **Section Reference:** [Where this information was found]\n• **Context:** [Additional context if needed]\n\n## ⚠️ **Important Notes**\n• [Any limitations, caveats, or important considerations]\n• [Whether content was truncated and might affect completeness]\n\n## 💡 **Summary**\n[Brief summary if the response was complex]\n\n**Remember:** Format your response to be highly readable with proper structure, bullet points, tables, and clear organization. Base your answer ONLY on the content provided.",
"options": {
"systemMessage": "You are a helpful document analysis assistant that provides clear, well-formatted answers based on document content."
},
"promptType": "define"
},
"typeVersion": 2.1
},
{
"id": "890ac685-4897-46aa-9e1a-e96566891b81",
"name": "工作流信息",
"type": "n8n-nodes-base.stickyNote",
"position": [
192,
-352
],
"parameters": {
"color": 4,
"width": 416,
"height": 528,
"content": "## 混合文档和网页分析器"
},
"typeVersion": 1
},
{
"id": "149ed301-6176-4f63-9890-7b69aaf8b264",
"name": "流水线信息",
"type": "n8n-nodes-base.stickyNote",
"position": [
672,
-160
],
"parameters": {
"color": 5,
"width": 500,
"height": 216,
"content": "## 简化的处理流程"
},
"typeVersion": 1
}
],
"active": false,
"pinData": {},
"settings": {
"executionOrder": "v1"
},
"versionId": "0b3b020f-958d-4e8c-acc8-5648076da1a5",
"connections": {
"c7024fab-3b1c-4892-abab-152a91c94b53": {
"main": [
[
{
"node": "bdb7e2fd-1441-4cac-a66e-dc20d8f04aeb",
"type": "main",
"index": 0
}
]
]
},
"8ad92e8a-54e5-4896-84aa-a054abc5c7f3": {
"main": [
[
{
"node": "afd1bad4-bc51-4f3d-8099-98dd9de1aac4",
"type": "main",
"index": 0
}
]
]
},
"8ec9d857-a965-47e0-a367-3172a1056232": {
"main": [
[
{
"node": "aaaefabe-e8fe-45b0-b841-109afc0049b3",
"type": "main",
"index": 0
}
]
]
},
"bdb7e2fd-1441-4cac-a66e-dc20d8f04aeb": {
"main": [
[
{
"node": "053e4752-6062-4641-8119-3d7aacbf6af1",
"type": "main",
"index": 0
}
]
]
},
"afd1bad4-bc51-4f3d-8099-98dd9de1aac4": {
"main": [
[
{
"node": "9a96077e-2158-42c0-853f-3dd43188c28d",
"type": "main",
"index": 0
}
]
]
},
"9a96077e-2158-42c0-853f-3dd43188c28d": {
"main": [
[
{
"node": "053e4752-6062-4641-8119-3d7aacbf6af1",
"type": "main",
"index": 0
}
]
]
},
"ce10391d-1e7b-4a68-9708-690961633dd1": {
"ai_languageModel": [
[
{
"node": "62a54e68-13ae-468c-acf0-b164138a19c0",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"053e4752-6062-4641-8119-3d7aacbf6af1": {
"main": [
[
{
"node": "62a54e68-13ae-468c-acf0-b164138a19c0",
"type": "main",
"index": 0
}
]
]
},
"aaaefabe-e8fe-45b0-b841-109afc0049b3": {
"main": [
[
{
"node": "8ad92e8a-54e5-4896-84aa-a054abc5c7f3",
"type": "main",
"index": 0
},
{
"node": "c7024fab-3b1c-4892-abab-152a91c94b53",
"type": "main",
"index": 0
}
]
]
}
}
}常见问题
如何使用这个工作流？
复制上方的 JSON 配置代码，在您的 n8n 实例中创建新工作流并选择「从 JSON 导入」，粘贴配置后根据需要修改凭证设置即可。
这个工作流适合什么场景？
中级
需要付费吗？
本工作流完全免费，您可以直接导入使用。但请注意，工作流中使用的第三方服务（如 OpenAI API）可能需要您自行付费。
相关工作流推荐
✨🩷自动化社交媒体内容发布工厂 + 系统提示组合
基于动态系统提示和GPT-4o的AI驱动多平台社交媒体内容工厂
If
Set
Code
+
If
Set
Code
100 节点Amit Mehta
内容创作
在可视化参考库中探索n8n节点
在可视化参考库中探索n8n节点
If
Ftp
Set
+
If
Ftp
Set
113 节点I versus AI
其他
✨🩷自动化社交媒体内容发布工厂 + 系统提示组合
使用GPT-4o为6个平台生成平台优化的社交媒体内容
If
Set
Code
+
If
Set
Code
100 节点Luan Correia
社交媒体
上下文混合RAG AI文案
Google Drive到Supabase上下文向量数据库同步用于RAG应用
If
Set
Code
+
If
Set
Code
76 节点Michael Taleb
AI RAG 检索增强
✨🩷 自动化社交媒体内容发布工厂 + 系统提示组合
✨🩷 自动化社交媒体内容发布工厂 + 系统提示组合
If
Set
Code
+
If
Set
Code
100 节点Joseph LePage
营销
与Supabase存储中文件对话的AI智能体
与Supabase存储中文件对话的AI智能体
If
Merge
Switch
+
If
Merge
Switch
33 节点Mark Shcherbakov
工程
工作流信息
难度等级
中级
节点数量12
分类-
节点类型9
作者
Aadarsh Jain
@aadarsh-jainAn enthusiast DevOps Engineer with 3+ years of experience in devops tools and automation.
外部链接
在 n8n.io 查看 →
分享此工作流