Customer Support Agent.json

{
  "name": "Customer Support Agent",
  "nodes": [
    {
      "parameters": {
        "workflowInputs": {
          "values": [
            {
              "name": "query"
            },
            {
              "name": "session_id"
            }
          ]
        }
      },
      "type": "n8n-nodes-base.executeWorkflowTrigger",
      "typeVersion": 1.1,
      "position": [
        0,
        0
      ],
      "id": "b94384af-5c09-49f7-8abe-9d5b1d1e8383",
      "name": "When Executed by Another Workflow"
    },
    {
      "parameters": {
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
      "typeVersion": 1,
      "position": [
        80,
        208
      ],
      "id": "de9d7d2a-353a-4c10-bbb5-fa2c73ff2c29",
      "name": "Google Gemini Chat Model",
      "credentials": {
        "googlePalmApi": {
          "id": "FCueVwq1thZXfmbB",
          "name": "demo_key"
        }
      }
    },
    {
      "parameters": {
        "mode": "retrieve-as-tool",
        "toolDescription": "Call this tool to Access the database.",
        "tableName": {
          "__rl": true,
          "value": "documents",
          "mode": "list",
          "cachedResultName": "documents"
        },
        "topK": 20,
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.vectorStoreSupabase",
      "typeVersion": 1.3,
      "position": [
        368,
        144
      ],
      "id": "60d07ff3-8a39-4e9d-a272-91414bc4cef3",
      "name": "Knowledge Base",
      "credentials": {
        "supabaseApi": {
          "id": "BF3uc6E3Vsqrb5H2",
          "name": "RAG System"
        }
      }
    },
    {
      "parameters": {},
      "type": "@n8n/n8n-nodes-langchain.embeddingsGoogleGemini",
      "typeVersion": 1,
      "position": [
        368,
        320
      ],
      "id": "82ed7237-b04d-4e1e-8a31-573baca9b104",
      "name": "Embeddings Google Gemini",
      "credentials": {
        "googlePalmApi": {
          "id": "FCueVwq1thZXfmbB",
          "name": "demo_key"
        }
      }
    },
    {
      "parameters": {
        "sessionIdType": "customKey",
        "sessionKey": "={{ $fromAI('session_id', 'the session id of the memory buffer') }}"
      },
      "type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
      "typeVersion": 1.3,
      "position": [
        224,
        208
      ],
      "id": "5b9d9ef6-3a6e-492f-b1ed-57cf5d2bfc0f",
      "name": "Simple Memory"
    },
    {
      "parameters": {
        "assignments": {
          "assignments": [
            {
              "id": "81e42018-dd7b-4187-a0a6-5abd20b7012b",
              "name": "output",
              "value": "={{ $json.output }}",
              "type": "string"
            }
          ]
        },
        "options": {}
      },
      "type": "n8n-nodes-base.set",
      "typeVersion": 3.4,
      "position": [
        560,
        0
      ],
      "id": "4327b27e-b32a-4881-85d9-bc8fe8e1fdba",
      "name": "Edit Fields"
    },
    {
      "parameters": {
        "content": "",
        "height": 688,
        "width": 912
      },
      "type": "n8n-nodes-base.stickyNote",
      "typeVersion": 1,
      "position": [
        -80,
        -160
      ],
      "id": "fbb4af90-fef1-425e-8e32-d7eb5375aaa8",
      "name": "Sticky Note"
    },
    {
      "parameters": {
        "content": "# Customer Support Agent",
        "height": 96,
        "width": 448
      },
      "type": "n8n-nodes-base.stickyNote",
      "typeVersion": 1,
      "position": [
        144,
        -160
      ],
      "id": "88ec3d8e-2b6c-45f2-bca2-bf7ad7071465",
      "name": "Sticky Note1"
    },
    {
      "parameters": {
        "promptType": "define",
        "text": "={{ $json.query }}",
        "options": {
          "systemMessage": "=# Overview  \nYou are a Retrieval-Augmented Generation (RAG) expert AI agent designed to answer user queries with high accuracy by retrieving information from authoritative sources within the Knowledge Base (KB).  \n\n## Context  \n- Your primary responsibility is to provide factually correct, concise, and well-sourced answers.  \n- All answers must be derived from the Knowledge Base, with explicit citation of MoF policies where relevant.  \n- You must avoid ambiguity, speculation, and unnecessary elaboration.  \n- You also have access to a memory tool containing past interactions. If a user query matches a previously answered question, you must retrieve the answer from memory instead of re-querying the Knowledge Base.  \n\n## Instructions  \n1. First, check the memory tool to determine if the current query has already been answered in a past interaction.  \n   - If yes, return the stored answer directly from memory for faster delivery.  \n   - If no, proceed to query the Knowledge Base.  \n2. Always query the Knowledge Base as many times as needed to ensure accuracy and completeness when no memory match exists.  \n3. Do not ask the user follow-up or clarifying questions. Answer directly, even if the query is ambiguous.  \n4. Provide responses in a concise, clear format that always ends with a **Citation** line.  \n5. Explicitly cite the Knowledge Base source(s) using the format:  \n   - Citation: [MoF Policy ###: Policy Title]  \n6. If the answer cannot be determined from the Knowledge Base, state:  \n   \"The information required to answer this question is not available in the provided knowledge bases.\"  \n\n## Tools  \n- Knowledge Base (KB) containing MoF policies and regulations  \n- Memory tool containing past interactions  \n\n## Examples  \n- Input: \"What are the requirements for Zone A?\"  \n- Output:  \n  \"Yes—Zone A requires submission ≥3 days before activation, allows up to 100 participants, and the nonprofit status and non-sensitive data comply with Zone A restrictions.  \n\n  Citation: [MoF Policy 001: Permit Eligibility by Zone]\"  \n\n- Input: \"Can I register an event in Zone C with 500 participants?\"  \n- Output:  \n  \"No—Zone C permits a maximum of 300 participants, therefore 500 participants exceeds the allowed limit.  \n\n  Citation: [MoF Policy 004: Zone Capacity Regulations]\"  \n\n- Input: \"What are the rules for Zone X?\"  \n- Output:  \n  \"The information required to answer this question is not available in the provided knowledge bases.\"  \n\n## SOP (Standard Operating Procedure)  \n1. Receive the user query.  \n2. Check the memory tool for a matching past interaction.  \n   - If a match exists, return the stored answer directly.  \n   - If no match exists, proceed to Step 3.  \n3. Search the Knowledge Base thoroughly for relevant MoF policies.  \n4. Retrieve authoritative information from the correct policy.  \n5. Construct a concise factual response in the approved output format.  \n6. Always end the answer with the **Citation line** referencing the specific MoF policy.  \n7. If no answer is available, return the fallback statement.  \n\n## Final Notes  \n- Always prioritize factual accuracy and proper MoF policy citation.  \n- Never speculate or provide answers unsupported by the Knowledge Base.  \n- Always follow the exact output format for consistency.  \n- Use memory first for efficiency; use KB only when memory does not contain the answer.  \n"
        }
      },
      "type": "@n8n/n8n-nodes-langchain.agent",
      "typeVersion": 2.2,
      "position": [
        208,
        0
      ],
      "id": "d7aeef89-8417-4bc6-9ac0-56d15cd8482b",
      "name": "Customer Support Agent"
    }
  ],
  "pinData": {
    "When Executed by Another Workflow": [
      {
        "json": {
          "query": "Can you summarize the policy number 1"
        }
      }
    ]
  },
  "connections": {
    "When Executed by Another Workflow": {
      "main": [
        [
          {
            "node": "Customer Support Agent",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Google Gemini Chat Model": {
      "ai_languageModel": [
        [
          {
            "node": "Customer Support Agent",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "Knowledge Base": {
      "ai_tool": [
        [
          {
            "node": "Customer Support Agent",
            "type": "ai_tool",
            "index": 0
          }
        ]
      ]
    },
    "Embeddings Google Gemini": {
      "ai_embedding": [
        [
          {
            "node": "Knowledge Base",
            "type": "ai_embedding",
            "index": 0
          }
        ]
      ]
    },
    "Simple Memory": {
      "ai_memory": [
        [
          {
            "node": "Customer Support Agent",
            "type": "ai_memory",
            "index": 0
          }
        ]
      ]
    },
    "Customer Support Agent": {
      "main": [
        [
          {
            "node": "Edit Fields",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  },
  "active": false,
  "settings": {
    "executionOrder": "v1"
  },
  "versionId": "2924e4f5-177a-4c73-b503-8d29372dee6d",
  "meta": {
    "templateCredsSetupCompleted": true,
    "instanceId": "9d139cc8817edbfe921ebb49cf470f37a38acccf02e4b0f7ecbfe97b65e3e56c"
  },
  "id": "XDGzcFdBP7ib9jFu",
  "tags": []
}