提示生成器
高级
这是一个Engineering, AI领域的自动化工作流，包含 21 个节点。主要使用 Form, Merge, SplitOut, FormTrigger, SplitInBatches 等节点，结合人工智能技术实现智能自动化。 AI提示生成器工作流
前置要求
- •Google Gemini API Key
使用的节点 (21)
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": "sCy7Dz1t2CXusiPB",
"meta": {
"instanceId": "2cdd9c593c5820bad56f93ddf9a9bb70d8c9e27fcdca4b21793502db1aae1d9d",
"templateCredsSetupCompleted": true
},
"name": "提示生成器",
"tags": [],
"nodes": [
{
"id": "4196a5c9-1217-4216-a3dc-ca752fc7e728",
"name": "表单提交时",
"type": "n8n-nodes-base.formTrigger",
"position": [
-1140,
-20
],
"webhookId": "d4dc2007-3c6d-4a26-b1f0-b5fbe6c8319c",
"parameters": {
"options": {
"customCss": "/* N8N Landing Page CSS - Correct Selectors Based on HTML */\n\n/* Style the main form card */\n.card {\n\tposition: relative;\n\tmax-width: 500px;\n\tmargin: 20px auto;\n\tpadding: 20px;\n\tbackground: white !important;\n\tborder-radius: 12px;\n\tbox-shadow: 0 10px 30px rgba(0, 0, 0, 0.1);\n\tfont-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;\n}\n\n/* Add custom header with gradient */\n.card:before {\n\tcontent: '';\n\tdisplay: block;\n\theight: 50px;\n\tbackground: linear-gradient(135deg, #6366f1 0%, #8b5cf6 100%);\n\tposition: relative;\n\tmargin: -20px -20px 30px -20px;\n\tborder-radius: 12px 12px 0 0;\n}\n\n/* Style the form header */\n.form-header {\n\ttext-align: center !important;\n\tcolor: #1f2937 !important;\n\tfont-weight: 700 !important;\n\tmargin: 80px 0 15px 0 !important;\n\tfont-size: 2.2em !important;\n\tposition: relative;\n\tz-index: 5;\n}\n\n/* Style any h1, h2 elements in the form */\n.card h1,\n.card h2 {\n\ttext-align: center !important;\n\tcolor: #1f2937 !important;\n\tfont-weight: 700 !important;\n\tmargin: 80px 0 15px 0 !important;\n\tfont-size: 2.2em !important;\n\tposition: relative;\n\tz-index: 5;\n}\n\n/* Style paragraphs and description text */\n.card p {\n\ttext-align: center !important;\n\tcolor: #6b7280 !important;\n\tfont-size: 1.1em !important;\n\tmargin-bottom: 30px !important;\n\tline-height: 1.5 !important;\n\tposition: relative;\n\tz-index: 5;\n}\n\n/* Style the inputs wrapper */\n.inputs-wrapper {\n\tposition: relative;\n\tz-index: 5;\n}\n\n/* Style all input fields */\n.card input[type=\"text\"],\n.card input[type=\"email\"],\n.card input[type=\"tel\"],\n.card input[type=\"number\"],\n.card textarea,\n.card select {\n\twidth: 100% !important;\n\tpadding: 12px 16px !important;\n\tborder: 2px solid #e5e7eb !important;\n\tborder-radius: 8px !important;\n\tfont-size: 16px !important;\n\ttransition: border-color 0.3s ease !important;\n\tmargin-bottom: 15px !important;\n\tbox-sizing: border-box !important;\n}\n\n.card input[type=\"text\"]:focus,\n.card input[type=\"email\"]:focus,\n.card input[type=\"tel\"]:focus,\n.card input[type=\"number\"]:focus,\n.card textarea:focus,\n.card select:focus {\n\toutline: none !important;\n\tborder-color: #6366f1 !important;\n\tbox-shadow: 0 0 0 3px rgba(99, 102, 241, 0.1) !important;\n}\n\n/* Style the submit button */\n.card button[type=\"submit\"],\n.card input[type=\"submit\"] {\n\tbackground: linear-gradient(135deg, #ff6b6b 0%, #ff8e8e 100%) !important;\n\tcolor: white !important;\n\tborder: none !important;\n\tpadding: 15px 40px !important;\n\tborder-radius: 10px !important;\n\tfont-size: 18px !important;\n\tfont-weight: 600 !important;\n\tcursor: pointer !important;\n\twidth: 100% !important;\n\tmargin-top: 20px !important;\n\ttransition: all 0.3s ease !important;\n\tbox-shadow: 0 4px 15px rgba(255, 107, 107, 0.3) !important;\n\tposition: relative;\n\tz-index: 5;\n}\n\n.card button[type=\"submit\"]:hover,\n.card input[type=\"submit\"]:hover {\n\ttransform: translateY(-2px) !important;\n\tbox-shadow: 0 6px 20px rgba(255, 107, 107, 0.4) !important;\n}\n\n/* Style the body to complement the form */\nbody.vsc-initialized {\n\tbackground: #f8fafc !important;\n\tfont-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif !important;\n\tmargin: 0 !important;\n\tpadding: 20px !important;\n}\n\n/* Style the container div */\n.container {\n\tmax-width: 600px !important;\n\tmargin: 0 auto !important;\n}\n\n/* Mobile responsive */\n@media (max-width: 768px) {\n\t.card {\n\t\tmargin: 10px !important;\n\t\tpadding: 15px !important;\n\t}\n\t\n\t.card:before {\n\t\theight: 50px !important;\n\t\tmargin: -15px -15px 20px -15px !important;\n\t}\n\t\n\t.card:after {\n\t\ttop: 35px !important;\n\t\tfont-size: 35px !important;\n\t}\n\t\n\t.form-header,\n\t.card h1,\n\t.card h2 {\n\t\tfont-size: 1.8em !important;\n\t\tmargin: 60px 0 15px 0 !important;\n\t}\n}",
"buttonLabel": "Let's Generate",
"appendAttribution": false
},
"formTitle": "🚀 AI Prompt Generator",
"formDescription": "Create powerful prompts for your AI tools"
},
"typeVersion": 2.2
},
{
"id": "bb9f2e02-a01e-4508-8ef0-37a9b9edd1ba",
"name": "Google Gemini聊天模型",
"type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
"position": [
-680,
180
],
"parameters": {
"options": {},
"modelName": "models/gemini-2.0-flash"
},
"credentials": {
"googlePalmApi": {
"id": "9BZpt0x33yMhJWqO",
"name": "Google Gemini(PaLM) Api account 2"
}
},
"typeVersion": 1
},
{
"id": "1fa60776-44ca-4478-b3bd-28b6c8fb8355",
"name": "结构化输出解析器",
"type": "@n8n/n8n-nodes-langchain.outputParserStructured",
"position": [
-520,
180
],
"parameters": {
"jsonSchemaExample": "[\n {\n \"fieldLabel\": \"Label for the question to ask\",\n \"placeholder\": \"Short hint to guide the user’s answer\",\n \"requiredField\": true,\n \"fieldType\": \"text\"\n }\n]"
},
"typeVersion": 1.2
},
{
"id": "2ad0f046-36ff-4864-af2e-339763109a9c",
"name": "基础问题",
"type": "n8n-nodes-base.form",
"position": [
-920,
-20
],
"webhookId": "cdf6ec24-43fe-4a6b-908a-9d17d3e0feec",
"parameters": {
"options": {
"customCss": "/* Apply to all n8n form pages */\n.card {\n\tposition: relative;\n\tmax-width: 500px;\n\tmargin: 20px auto;\n\tpadding: 20px;\n\tbackground: white !important;\n\tborder-radius: 12px;\n\tbox-shadow: 0 10px 30px rgba(0, 0, 0, 0.1);\n\tfont-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;\n}\n\n.card:before {\n\tcontent: '';\n\tdisplay: block;\n\theight: 50px;\n\tbackground: linear-gradient(135deg, #6366f1 0%, #8b5cf6 100%);\n\tposition: relative;\n\tmargin: -20px -20px 30px -20px;\n\tborder-radius: 12px 12px 0 0;\n}\n\n.card h1, .card h2 {\n\ttext-align: center !important;\n\tcolor: #1f2937 !important;\n\tfont-weight: 700 !important;\n\tmargin: 15px 0 !important;\n\tfont-size: 2em !important;\n}\n\n.card p {\n\tfont-size: 1.1em !important;\n\tline-height: 1.5 !important;\n}\n\n.inputs-wrapper {\n\tposition: relative;\n}\n\n.form-group {\n margin-bottom: 15px;\n}\n\n.card input, .card textarea, .card select {\n\twidth: 100% !important;\n\tpadding: 12px 16px !important;\n\tborder: 2px solid #e5e7eb !important;\n\tborder-radius: 8px !important;\n\tfont-size: 16px !important;\n\tbox-sizing: border-box !important;\n\ttransition: border-color 0.3s ease !important;\n}\n\n.card input:focus, .card textarea:focus, .card select:focus {\n\tborder-color: #6366f1 !important;\n\tbox-shadow: 0 0 0 3px rgba(99, 102, 241, 0.1) !important;\n\toutline: none !important;\n}\n\n.card button[type=\"submit\"] {\n\tbackground: linear-gradient(135deg, #16a34a 0%, #22c55e 100%) !important;\n\tcolor: white !important;\n\tpadding: 15px 40px !important;\n\tborder-radius: 10px !important;\n\tfont-size: 18px !important;\n\tfont-weight: 600 !important;\n\tcursor: pointer !important;\n\twidth: 100% !important;\n\tbox-shadow: 0 4px 15px rgba(34, 197, 94, 0.3) !important;\n\ttransition: all 0.3s ease !important;\n}\n\n.card button[type=\"submit\"]:hover {\n\ttransform: translateY(-2px) !important;\n\tbox-shadow: 0 6px 20px rgba(34, 197, 94, 0.4) !important;\n}\n\n.error-hidden {\n display: block;\n position: relative;\n color: var(--color-error);\n text-align: left;\n font-size: var(--font-size-error);\n font-weight: 400;\n visibility: hidden;\n padding-top: 0;\n padding-bottom: 0;\n}\n\n@media (max-width: 768px) {\n\t.card {\n\t\tmargin: 10px !important;\n\t\tpadding: 15px !important;\n\t}\n\t.card h1, .card h2 {\n\t\tfont-size: 1.6em !important;\n\t}\n .card:before {\n margin: -20px -15px 30px -15px;\n }\n}\n",
"formTitle": "Enrich Prompt",
"buttonLabel": "Answer"
},
"formFields": {
"values": [
{
"fieldType": "textarea",
"fieldLabel": "What do you want to build ?",
"placeholder": "i.e. A B2B proposal generator for content marketing agency",
"requiredField": true
},
{
"fieldType": "textarea",
"fieldLabel": "Tools I can access (N/A if no tools)",
"placeholder": "i.e. Web Search, Email Threads, Google Sheets",
"requiredField": true
},
{
"fieldType": "textarea",
"fieldLabel": "What Input can be expected ?",
"placeholder": "i.e. The customer data like name, company, problems you can solve.",
"requiredField": true
},
{
"fieldType": "textarea",
"fieldLabel": "What output do you expect ?",
"placeholder": "i.e. A proposal for the company that is specific to the customer",
"requiredField": true
}
]
}
},
"typeVersion": 1
},
{
"id": "f7e1522a-3951-47fb-9e5a-5730ff3cf055",
"name": "循环问题",
"type": "n8n-nodes-base.splitInBatches",
"position": [
0,
60
],
"parameters": {
"options": {}
},
"executeOnce": false,
"typeVersion": 3
},
{
"id": "e2b5d3c4-cb4c-4f90-93a1-10784bd8a38e",
"name": "相关问题",
"type": "n8n-nodes-base.form",
"position": [
260,
80
],
"webhookId": "8d220b4e-1c9c-4f52-8d1f-0e5f396f7745",
"parameters": {
"options": {
"customCss": "/* Apply to all n8n form pages */\n.card {\n\tposition: relative;\n\tmax-width: 500px;\n\tmargin: 20px auto;\n\tpadding: 20px;\n\tbackground: white !important;\n\tborder-radius: 12px;\n\tbox-shadow: 0 10px 30px rgba(0, 0, 0, 0.1);\n\tfont-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;\n}\n\n.card:before {\n\tcontent: '';\n\tdisplay: block;\n\theight: 50px;\n\tbackground: linear-gradient(135deg, #6366f1 0%, #8b5cf6 100%);\n\tposition: relative;\n\tmargin: -20px -20px 30px -20px;\n\tborder-radius: 12px 12px 0 0;\n}\n\n.card h1, .card h2 {\n\ttext-align: center !important;\n\tcolor: #1f2937 !important;\n\tfont-weight: 700 !important;\n\tmargin: 15px 0 !important;\n\tfont-size: 2em !important;\n}\n\n.card p {\n\tfont-size: 1.1em !important;\n\tline-height: 1.5 !important;\n}\n\n.inputs-wrapper {\n\tposition: relative;\n}\n\n.form-group {\n margin-bottom: 15px;\n}\n\n.card input, .card textarea, .card select {\n\twidth: 100% !important;\n\tpadding: 12px 16px !important;\n\tborder: 2px solid #e5e7eb !important;\n\tborder-radius: 8px !important;\n\tfont-size: 16px !important;\n\tbox-sizing: border-box !important;\n\ttransition: border-color 0.3s ease !important;\n}\n\n.card input:focus, .card textarea:focus, .card select:focus {\n\tborder-color: #6366f1 !important;\n\tbox-shadow: 0 0 0 3px rgba(99, 102, 241, 0.1) !important;\n\toutline: none !important;\n}\n\n.card button[type=\"submit\"] {\n\tbackground: linear-gradient(135deg, #16a34a 0%, #22c55e 100%) !important;\n\tcolor: white !important;\n\tpadding: 15px 40px !important;\n\tborder-radius: 10px !important;\n\tfont-size: 18px !important;\n\tfont-weight: 600 !important;\n\tcursor: pointer !important;\n\twidth: 100% !important;\n\tbox-shadow: 0 4px 15px rgba(34, 197, 94, 0.3) !important;\n\ttransition: all 0.3s ease !important;\n}\n\n.card button[type=\"submit\"]:hover {\n\ttransform: translateY(-2px) !important;\n\tbox-shadow: 0 6px 20px rgba(34, 197, 94, 0.4) !important;\n}\n\n.error-hidden {\n display: block;\n position: relative;\n color: var(--color-error);\n text-align: left;\n font-size: var(--font-size-error);\n font-weight: 400;\n visibility: hidden;\n padding-top: 0;\n padding-bottom: 0;\n}\n\n@media (max-width: 768px) {\n\t.card {\n\t\tmargin: 10px !important;\n\t\tpadding: 15px !important;\n\t}\n\t.card h1, .card h2 {\n\t\tfont-size: 1.6em !important;\n\t}\n .card:before {\n margin: -20px -15px 30px -15px;\n }\n}\n",
"formTitle": "Questions to understand",
"buttonLabel": "Answer"
},
"defineForm": "json",
"jsonOutput": "=[{{ $('LoopQuestions').item.json }}]"
},
"typeVersion": 1
},
{
"id": "1618be1e-222c-45d5-a85e-d76878971f60",
"name": "合并用户意图",
"type": "n8n-nodes-base.merge",
"position": [
500,
-20
],
"parameters": {},
"typeVersion": 3.1
},
{
"id": "cbd1fbb5-a530-4695-9f7a-d3c1fe394bb9",
"name": "提示生成器",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
740,
-20
],
"parameters": {
"text": "={{ $('MergeUserIntent').item.json }}",
"messages": {
"messageValues": [
{
"message": "<role>\nYou are an expert AI prompt generator designed to turn semi-structured user inputs into effective and well-formatted prompts for downstream AI agents. Your goal is to craft high-quality prompts that follow strict formatting and enable the target AI agent to perform its task optimally. And to do that you'll be receiving the user's input like what they want to build, which tool AI can use, what input the ai need to expect and what output the user want from their ai agent. Also you'll be receiving some specific constraints for the ai agent to be taken care along with some clarifying questions.\n</role>\n\n<inputs>\nYou will receive a JSON array that includes:\n- The user's primary intent (e.g., what they want to build)\n- What tools they plan to use (or \"N/A\")\n- What input the tool will receive\n- What output they expect\n- Any constraints (e.g., token limits, length)\n- Up to 3 additional form answers for clarifying context (e.g., tone, style, audience)\n\nExample input:\n[\n {\n \"What do you want to build ?\": \"Video to LinkedIn post\",\n \"Tools I can access (N/A if no tools)\": \"N/A\",\n \"What Input can be expected ?\": \"video\",\n \"What output do you expect ?\": \"LinkedIn Post\",\n \"submittedAt\": \"2025-06-12T00:47:05.645+05:30\",\n \"formMode\": \"test\"\n },\n {\n \"Preferred tone or style for the LinkedIn post\": \"witty\",\n \"submittedAt\": \"2025-06-12T01:15:11.608+05:30\",\n \"formMode\": \"test\"\n },\n {\n \"Should the post summarize the video or extract quotes?\": \"both\",\n \"submittedAt\": \"2025-06-12T01:15:14.925+05:30\",\n \"formMode\": \"test\"\n },\n {\n \"Who is the target audience for the LinkedIn post?\": \"general network\",\n \"submittedAt\": \"2025-06-12T01:15:20.045+05:30\",\n \"formMode\": \"test\"\n }\n]\n</inputs>\n\n<tools>\nYou do not have access to APIs, databases, or external documents. Use only the input provided.\n</tools>\n\n<instructions>\n1. Parse and extract all user-provided values from the JSON array.\n2. Understand the user’s goal, expected inputs and outputs, and key contextual details.\n3. Using that information, generate a final prompt in strictly the following format (Critics) where r and c will change the position r first and then c:\n\n- **<constraints>**: Any boundaries (length, style, tone, technical limits)\n- **<role>**: What the downstream AI agent’s expertise and purpose should be \n- **<inputs>**: What inputs it will receive, in what format, and example values \n- **<tools>**: Any tools or data sources it can or cannot access \n- **<instructions>**: Step-by-step guidance for the agent to perform its task \n- **<conclusions>**: What kind of output it should generate, with format or examples \n\n4. Ensure the final prompt is useful for a general-purpose LLM, without ambiguity.\n</instructions>\n\n<constraints>\n- Always include all 6 sections: Role, Inputs, Tools, Instructions, Constraints, Conclusions\n- Be specific and concise, avoid generic phrasing\n- Do not include timestamps or metadata like \"formMode\" or \"submittedAt\"\n- Output must be readable by a prompt engineer or AI agent directly\n</constraints>\n\n<conclusions>\nYour output will be a complete structured prompt, ready to be used by a downstream LLM for task execution.\n\nExample output:\n<role>\nYou are an AI assistant that transforms short videos into compelling LinkedIn posts. Your job is to summarise the video, extract powerful quotes, and write in a witty tone suitable for a general professional audience.\n</role>\n\n<inputs>\nYou will receive:\n- A video (maximum 5 minutes)\n- Preferences for tone: witty\n- Summary style: both (summarize and extract quotes)\n- Target audience: general network\n</inputs>\n\n<tools>\nYou do not have access to external APIs or web search. You can assume the video has already been transcribed for processing.\n</tools>\n\n<instructions>\n1. Parse the video transcript and understand its core message.\n2. Identify key insights and at least one memorable quote.\n3. Write a LinkedIn post that begins with a strong hook, reflects a witty tone, and engages the general network.\n4. Balance insight and personality while staying platform-native.\n</instructions>\n\n<constraints>\n- Keep the post under 3000 characters\n- Avoid technical jargon unless it serves the audience\n- No generic intros or robotic tone\n</constraints>\n\n<conclusions>\nReturn a single, polished LinkedIn post. Example:\n\n---\n“Most products don’t need onboarding. They need rethinking.”\n\nJust watched a great clip on UX that reminded me how design isn’t about making things easier — it’s about making them unnecessary.\n\n#UX #DesignThinking\n---\n</conclusions>\n\n<output>\nReturn strictly only a JSON object like this:\n{\n \"prompt\": \"<your generated prompt string>\"\n}\nDo NOT wrap it inside code blocks or markdown.\n</output>"
}
]
},
"promptType": "define",
"hasOutputParser": true
},
"executeOnce": true,
"typeVersion": 1.6
},
{
"id": "869e9a51-0e04-4bdc-8cf3-b35573b2953b",
"name": "Google Gemini 聊天模型1",
"type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
"position": [
760,
360
],
"parameters": {
"options": {},
"modelName": "models/gemini-2.0-flash"
},
"credentials": {
"googlePalmApi": {
"id": "9BZpt0x33yMhJWqO",
"name": "Google Gemini(PaLM) Api account 2"
}
},
"typeVersion": 1
},
{
"id": "cef87b06-dfd2-49ee-a294-75d3eabee58d",
"name": "便签",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1220,
380
],
"parameters": {
"width": 900,
"height": 300,
"content": "# 提示工程"
},
"typeVersion": 1
},
{
"id": "cb8ee56d-7d41-4f37-a640-c8aea8588707",
"name": "相关问题AI",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"position": [
-700,
-20
],
"parameters": {
"text": "={{ $json }}",
"messages": {
"messageValues": [
{
"message": "<role>\nYou are an expert LLM-based prompt refinement agent. Your goal is to analyze vague or semi-structured input from a user who wants to build an automation or AI-based tool. Based on what they share, your task is to generate 3 **highly specific** questions that, if answered, would help a prompt engineer craft a more tailored and effective prompt for the tool they wish to build. The questions you generate should be very specific to the ai agent the user wants to build, and completely relevant for the idea they want to build it for.\n</role>\n\n<input>\nYou will receive a JSON-like array with the following fields:\n- \"What do you want to build ?\" – Describes the general idea.\n- \"Tools I can access (N/A if no tools)\" – Lists any tools/services the user can integrate.\n- \"What Input can be expected ?\" – The user input or data source for the tool.\n- \"What output do you expect ?\" – The final result the tool should generate.\n- \"Any Constraints for the tool (N/A if no constraints)\" – Character/length/performance/time restrictions.\n\nEach field is plain text. The values may vary in clarity and depth.\n</input>\n\n<tools>\nYou do not have access to external tools. Only the given input array and your own reasoning ability are available. You cannot fetch additional context.\n</tools>\n\n<instructions>\n1. Parse and understand the user's goal.\n2. Identify any ambiguity or key gaps in what’s provided.\n3. Generate **3 clear, concise follow-up questions** that would best clarify:\n - Context or tone expectations\n - Format or structure needed\n - Any target audience or usage specifics\n4. Explain **why** each question is important (in 1–2 sentences).\n5. Return your output in the strict JSON array format defined above.\n6. Identify missing but critical details that would affect how an AI agent should behave.\n7. For each missing detail, create one structured form field in the output format.\n8. Use fieldLabel for the question, placeholder to guide, and set requiredField appropriately.\n9. Optionally specify fieldType (e.g., textarea for long text, dropdown if fixed options are obvious).\n</instructions>\n\n<constraints>\n\nYour output must be in the following format:\n\n[\n {\n \"fieldLabel\": \"Label for the question to ask\",\n \"placeholder\": \"Short hint to guide the user’s answer\",\n \"requiredField\": true or false,\n \"fieldType\": \"number\" | \"email\" | \"textarea\" | \"dropdown\" (not optional; take default as 'textarea')\n }\n]\n\n- Ask no more than 3 questions.\n- Only ask questions that meaningfully impact how the tool would be built or how the AI would behave.\n- Your questions must be simple, not technical or abstract.\n- Do not repeat or rephrase questions already answered in the input.\n- Be practical — ask questions that a human prompt engineer would truly need to ask to make the result more useful.\n</constraints>\n\n<output>\n[\n {\n \"fieldLabel\": \"Preferred tone or style for the LinkedIn post\",\n \"placeholder\": \"e.g., professional, friendly, witty\",\n \"requiredField\": true,\n \"fieldType\": \"textarea\"\n },\n {\n \"fieldLabel\": \"Should the post summarize the video or extract quotes?\",\n \"placeholder\": \"e.g., summary only, key quotes, both\",\n \"requiredField\": true,\n \"fieldType\": \"textarea\"\n },\n {\n \"fieldLabel\": \"Who is the target audience for the LinkedIn post?\",\n \"placeholder\": \"e.g., hiring managers, founders, general network\",\n \"requiredField\": false,\n \"fieldType\": \"textarea\"\n }\n]\n</output>"
}
]
},
"promptType": "define",
"hasOutputParser": true
},
"typeVersion": 1.6
},
{
"id": "6e88dd2e-6aa3-46b1-ae4e-437fcdd7d909",
"name": "拆分问题",
"type": "n8n-nodes-base.splitOut",
"position": [
-280,
-20
],
"parameters": {
"options": {},
"fieldToSplitOut": "output"
},
"typeVersion": 1
},
{
"id": "d38f8848-8409-4ab5-a983-b927baa47585",
"name": "自动修复输出解析器",
"type": "@n8n/n8n-nodes-langchain.outputParserAutofixing",
"position": [
880,
140
],
"parameters": {
"options": {}
},
"typeVersion": 1
},
{
"id": "de385387-86c1-45a2-a750-4f4a9b70cb0c",
"name": "结构化输出解析器1",
"type": "@n8n/n8n-nodes-langchain.outputParserStructured",
"position": [
960,
360
],
"parameters": {
"jsonSchemaExample": "{\n\t\"prompt\": \"this is the prompt\"\n}"
},
"typeVersion": 1.2
},
{
"id": "9ba8e561-6052-4d5c-915b-ebd7eeb910ad",
"name": "便签1",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1220,
-160
],
"parameters": {
"width": 440,
"height": 320,
"content": "# 初始化并获取基础问题"
},
"typeVersion": 1
},
{
"id": "5730302d-6e4b-435c-8e51-43d389d3f065",
"name": "便签2",
"type": "n8n-nodes-base.stickyNote",
"position": [
-760,
-160
],
"parameters": {
"width": 380,
"height": 500,
"content": "# 生成相关问题"
},
"typeVersion": 1
},
{
"id": "f8f6640c-e97a-44b1-87bc-687783541309",
"name": "便签3",
"type": "n8n-nodes-base.stickyNote",
"position": [
-320,
-160
],
"parameters": {
"width": 960,
"height": 500,
"content": "# 向用户提问"
},
"typeVersion": 1
},
{
"id": "cc14b692-ccf5-4ad0-9f59-1ec4eee5adfa",
"name": "便签4",
"type": "n8n-nodes-base.stickyNote",
"position": [
660,
-160
],
"parameters": {
"width": 520,
"height": 680,
"content": "# 提示生成器系统"
},
"typeVersion": 1
},
{
"id": "b42d7917-d7b4-4427-8530-c541ea7c1cef",
"name": "便利贴5",
"type": "n8n-nodes-base.stickyNote",
"position": [
1220,
-160
],
"parameters": {
"width": 300,
"height": 320,
"content": "# 向用户发送提示"
},
"typeVersion": 1
},
{
"id": "3e44d786-2cb2-41d8-8568-d0c50dc98c55",
"name": "发送提示",
"type": "n8n-nodes-base.form",
"position": [
1280,
-20
],
"webhookId": "9a0af1f5-3257-486c-ad6e-068a7c0f93ab",
"parameters": {
"options": {
"customCss": ".card {\n\tposition: relative;\n\tmax-width: 500px;\n\tmargin: 20px auto;\n\tpadding: 20px;\n\tbackground: #fefefe !important;\n\tborder-radius: 12px;\n\tbox-shadow: 0 10px 30px rgba(0, 0, 0, 0.1);\n\tfont-family: 'Courier New', monospace;\n\tcolor: #1f2937;\n\tfont-size: 16px;\n\tline-height: 1.6;\n\twhite-space: pre-wrap;\n}\n\n.header p {\n text-align: left;\n}\n\n@media (max-width: 768px) {\n\t.card {\n\t\tmargin: 10px !important;\n\t\tpadding: 15px !important;\n\t\tfont-size: 14px;\n\t}\n}\n"
},
"operation": "completion",
"completionTitle": "🎉 Here's your custom prompt",
"completionMessage": "={{ $('PromptGenerator').item.json.output.prompt }}\n\n---------------------------\n\n📋 You can now copy and use it anywhere!"
},
"typeVersion": 1
},
{
"id": "176312e6-c5e6-4cba-ab4a-c8aeb5ee1722",
"name": "便签 6",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1220,
-300
],
"parameters": {
"color": 7,
"width": 440,
"height": 100,
"content": "# 🚀 AI提示生成器"
},
"typeVersion": 1
}
],
"active": true,
"pinData": {},
"settings": {
"callerPolicy": "workflowsFromSameOwner",
"executionOrder": "v1",
"executionTimeout": -1
},
"versionId": "6cdac3f9-1238-49d1-b6a9-01e2a3a4f273",
"connections": {
"BaseQuestions": {
"main": [
[
{
"node": "RelatedQuestionAI",
"type": "main",
"index": 0
},
{
"node": "MergeUserIntent",
"type": "main",
"index": 0
}
]
]
},
"LoopQuestions": {
"main": [
[
{
"node": "MergeUserIntent",
"type": "main",
"index": 1
}
],
[
{
"node": "RelevantQuestions",
"type": "main",
"index": 0
}
]
]
},
"SplitQuestions": {
"main": [
[
{
"node": "LoopQuestions",
"type": "main",
"index": 0
}
]
]
},
"MergeUserIntent": {
"main": [
[
{
"node": "PromptGenerator",
"type": "main",
"index": 0
}
]
]
},
"PromptGenerator": {
"main": [
[
{
"node": "SendingPrompt",
"type": "main",
"index": 0
}
]
]
},
"RelatedQuestionAI": {
"main": [
[
{
"node": "SplitQuestions",
"type": "main",
"index": 0
}
]
]
},
"RelevantQuestions": {
"main": [
[
{
"node": "LoopQuestions",
"type": "main",
"index": 0
}
]
]
},
"On form submission": {
"main": [
[
{
"node": "BaseQuestions",
"type": "main",
"index": 0
}
]
]
},
"Google Gemini Chat Model": {
"ai_languageModel": [
[
{
"node": "RelatedQuestionAI",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Structured Output Parser": {
"ai_outputParser": [
[
{
"node": "RelatedQuestionAI",
"type": "ai_outputParser",
"index": 0
}
]
]
},
"Auto-fixing Output Parser": {
"ai_outputParser": [
[
{
"node": "PromptGenerator",
"type": "ai_outputParser",
"index": 0
}
]
]
},
"Google Gemini Chat Model1": {
"ai_languageModel": [
[
{
"node": "PromptGenerator",
"type": "ai_languageModel",
"index": 0
},
{
"node": "Auto-fixing Output Parser",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Structured Output Parser1": {
"ai_outputParser": [
[
{
"node": "Auto-fixing Output Parser",
"type": "ai_outputParser",
"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？
高级 - 工程, 人工智能
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
使用 n8n、Apify 和 OpenAI o3 自托管 AI 深度研究代理
使用 n8n、Apify 和 OpenAI o3 自托管 AI 深度研究代理
If
Set
Code
+20
87 节点Jimleuk
人工智能
使用正则表达式和 AI 发现隐藏的网站 API 端点
使用正则表达式和人工智能发现隐藏的网站API端点
If
Set
Html
+19
58 节点Yulia
工程
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
+93
113 节点I versus AI
其他
Reddit MVP生成器
使用AI和Excel存储从Reddit自动生成MVP创业想法
Code
Merge
Reddit
+10
30 节点Jesse White
产品
通过Bright Data提取、总结和分析亚马逊产品降价情况
使用Bright Data和Google Gemini提取、总结和分析亚马逊降价信息
Set
Wait
Merge
+14
26 节点Ranjan Dailata
人工智能
（Duc）深度研究市场模板
集成PerplexityAI研究和OpenAI内容的多层级WordPress博客生成器
If
Set
Xml
+28
132 节点Daniel Ng
人工智能
