{
  "name": "SupportFlow Lite — Simple AI Support Bot (No RAG)",
  "nodes": [
    {
      "parameters": {
        "httpMethod": "POST",
        "path": "supportflow-lite",
        "responseMode": "responseNode",
        "options": {}
      },
      "id": "trigger-webhook",
      "name": "Receive Message",
      "type": "n8n-nodes-base.webhook",
      "typeVersion": 2,
      "position": [200, 400],
      "webhookId": "supportflow-lite"
    },
    {
      "parameters": {
        "jsCode": "const body = $input.first().json.body || $input.first().json;\nconst message = body.message || body.text || '';\nconst sessionId = body.session_id || require('crypto').randomUUID();\nconst customerName = body.customer_name || 'Customer';\n\nif (!message || message.trim().length < 2) {\n  throw new Error('Message is required.');\n}\n\nreturn [{ json: { message: message.trim(), sessionId, customerName, receivedAt: new Date().toISOString(), messageId: require('crypto').randomUUID() } }];"
      },
      "id": "validate-input",
      "name": "Validate Input",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [420, 400]
    },
    {
      "parameters": {
        "model": "gpt-4o-mini",
        "messages": {
          "values": [
            {
              "role": "system",
              "content": "=You are a helpful customer support assistant for {{ $env.SUPPORTFLOW_COMPANY_NAME || 'our company' }}.\n\nCompany info:\n{{ $env.SUPPORTFLOW_COMPANY_INFO || 'We are a technology company focused on providing excellent customer service.' }}\n\nRules:\n1. Be friendly, professional, and concise\n2. Answer based on the company info provided\n3. If you don't know the answer, say: 'I don't have that information. Let me connect you with our team at [support email].'\n4. Keep responses under 150 words\n5. Never make up product features, prices, or policies\n\nRespond in JSON: { \"answer\": \"...\", \"confidence\": 0.0-1.0, \"category\": \"faq|billing|technical|complaint|other\" }"
            },
            {
              "role": "user",
              "content": "=Customer {{$json.customerName}} asks:\n{{$json.message}}"
            }
          ]
        },
        "options": { "temperature": 0.3, "maxTokens": 500, "responseFormat": "json_object" }
      },
      "id": "ai-respond",
      "name": "AI: Generate Response",
      "type": "n8n-nodes-base.openAi",
      "typeVersion": 1.8,
      "position": [660, 400],
      "retryOnFail": true,
      "maxTries": 3,
      "waitBetweenTries": 2000
    },
    {
      "parameters": {
        "jsCode": "const original = $('Validate Input').first().json;\nconst aiRaw = $input.first().json;\nlet ai;\ntry {\n  ai = JSON.parse(aiRaw.message?.content || aiRaw.content || '{}');\n} catch (e) {\n  ai = { answer: 'Sorry, I had trouble processing that. Please try again.', confidence: 0, category: 'other' };\n}\n\nreturn [{ json: { ...original, answer: ai.answer || '', confidence: ai.confidence || 0, category: ai.category || 'other', respondedAt: new Date().toISOString() } }];"
      },
      "id": "merge-response",
      "name": "Merge Response",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [900, 400]
    },
    {
      "parameters": {
        "operation": "append",
        "documentId": { "__rl": true, "value": "={{ $env.SUPPORTFLOW_SHEET_ID }}", "mode": "id" },
        "sheetName": { "__rl": true, "value": "Lite Conversations", "mode": "name" },
        "columns": {
          "mappingMode": "defineBelow",
          "value": {
            "Message ID": "={{ $json.messageId }}",
            "Session ID": "={{ $json.sessionId }}",
            "Customer": "={{ $json.customerName }}",
            "Question": "={{ $json.message }}",
            "Answer": "={{ $json.answer }}",
            "Confidence": "={{ $json.confidence }}",
            "Category": "={{ $json.category }}",
            "Received": "={{ $json.receivedAt }}",
            "Responded": "={{ $json.respondedAt }}"
          }
        },
        "options": {}
      },
      "id": "log-sheets",
      "name": "Log to Sheets",
      "type": "n8n-nodes-base.googleSheets",
      "typeVersion": 4.5,
      "position": [1140, 300],
      "retryOnFail": true,
      "maxTries": 2
    },
    {
      "parameters": {
        "respondWith": "json",
        "responseBody": "={{ JSON.stringify({ success: true, messageId: $json.messageId, answer: $json.answer, confidence: $json.confidence, category: $json.category }) }}"
      },
      "id": "respond",
      "name": "Respond to Customer",
      "type": "n8n-nodes-base.respondToWebhook",
      "typeVersion": 1.1,
      "position": [1140, 500]
    },
    {
      "parameters": {
        "content": "## SupportFlow Lite — Simple AI Bot\n\nNo vector database needed.\nUses company info from environment variable.\n\nBest for: Small businesses, MVPs, testing.\nSetup: 2 minutes (just OpenAI + Sheets).\n\nUpgrade to Main workflow when you\nneed RAG-powered knowledge base.",
        "width": 350,
        "height": 220,
        "color": 6
      },
      "id": "note-overview",
      "name": "Sticky Note: Lite",
      "type": "n8n-nodes-base.stickyNote",
      "typeVersion": 1,
      "position": [120, 140]
    }
  ],
  "connections": {
    "Receive Message": {
      "main": [[{ "node": "Validate Input", "type": "main", "index": 0 }]]
    },
    "Validate Input": {
      "main": [[{ "node": "AI: Generate Response", "type": "main", "index": 0 }]]
    },
    "AI: Generate Response": {
      "main": [[{ "node": "Merge Response", "type": "main", "index": 0 }]]
    },
    "Merge Response": {
      "main": [[
        { "node": "Log to Sheets", "type": "main", "index": 0 },
        { "node": "Respond to Customer", "type": "main", "index": 0 }
      ]]
    }
  },
  "settings": { "executionOrder": "v1", "saveManualExecutions": true },
  "tags": [
    { "name": "SupportFlow", "id": "supportflow" },
    { "name": "AutoFlow", "id": "autoflow" },
    { "name": "Lite", "id": "lite" }
  ],
  "meta": { "instanceId": "supportflow-lite-v1" }
}