8
n8n 中文网amn8n.com

使用AI自动分类Gmail邮件并应用标签整理收件箱

中级

这是一个Other, AI领域的自动化工作流,包含 8 个节点。主要使用 Code, Gmail, Merge, Aggregate, GmailTrigger 等节点,结合人工智能技术实现智能自动化。 使用AI自动分类Gmail邮件并应用标签整理收件箱

前置要求
  • Google 账号和 Gmail API 凭证
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
{
  "meta": {
    "instanceId": "5a90cf25b26f3375705646dd979ac7968c2f7780e9cb14036112ef363426c72c",
    "templateCredsSetupCompleted": true
  },
  "nodes": [
    {
      "id": "37166e51-ffb6-43fe-8bd5-fe5eb1249366",
      "name": "OpenRouter聊天模型",
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenRouter",
      "position": [
        540,
        540
      ],
      "parameters": {
        "model": "google/gemini-2.5-flash-preview-05-20",
        "options": {}
      },
      "credentials": {
        "openRouterApi": {
          "id": "ogejgDjpnXZzGonu",
          "name": "OpenRouter account"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "067f6b55-dc4c-4757-b19e-2d539b8b2526",
      "name": "聚合 Gmail 邮件",
      "type": "n8n-nodes-base.aggregate",
      "position": [
        -360,
        220
      ],
      "parameters": {
        "options": {},
        "aggregate": "aggregateAllItemData"
      },
      "typeVersion": 1
    },
    {
      "id": "7807ff72-49f7-4a7b-9d75-24746d0dfdb7",
      "name": "合并标签与邮件",
      "type": "n8n-nodes-base.merge",
      "position": [
        240,
        320
      ],
      "parameters": {
        "mode": "combine",
        "options": {},
        "combineBy": "combineAll"
      },
      "typeVersion": 3.1
    },
    {
      "id": "8c51f04b-3415-4d53-8624-ce8d484145b9",
      "name": "使用 AI 分类邮件",
      "type": "@n8n/n8n-nodes-langchain.chainLlm",
      "position": [
        520,
        320
      ],
      "parameters": {
        "text": "=You are an email classification assistant. Your task is to analyze Gmail email snippets and categorize them into one of the following labels:\n\n1. To respond - Emails requiring a direct personal response from you (questions from individuals, personal conversations, NOT automated feedback requests)\n2. FYI - Emails requesting you to take action, reminders, or alerts about issues (add money requests, payment failures, action needed)\n3. Comment - Emails specifically asking for your feedback, review, or opinion in a professional/collaborative context\n4. Notification - Automated updates about completed or ongoing events (successful payments, transfers sent, deployments, shipments, security alerts)\n5. Meeting update - Calendar invites, meeting changes, or meeting-related communications\n6. Awaiting reply - Follow-up emails or reminders about your pending responses\n7. Actioned - Confirmations of actions YOU specifically took (not automated system actions)\n8. Marketing - Promotional content, sales emails, feature announcements, newsletters, or automated customer feedback surveys\n\nKey classification rules:\n- Automated status updates (deployments, transfers, payments processed) → Notification\n- Requests to add money or take action → FYI\n- Payment failure alerts → FYI\n- Security alert digests → Notification\n- Shipment updates → Notification\n- Direct Debit payment confirmations → Notification\n- Balance update confirmations → Notification\n- Customer satisfaction surveys or feedback requests → Marketing\n- Automated \"how was your experience\" emails → Marketing\n- Only use \"Actioned\" for confirmations of YOUR manual actions\n- Only use \"Marketing\" for promotional content, including feedback surveys\n\nInstructions:\n1. Read the email snippet carefully\n2. Note the sender (if provided) for context\n3. Choose the single most appropriate label based on the rules above\n4. IMPORTANT: Respond ONLY with the number and label (e.g., \"2. FYI\")\n5. Do not include any other text, explanation, or repeat the email content\n\nEmail From:\n{{ $json.From }}\nEmail To:\n{{ $json.To }}\nEmail subject:\n{{ $json.Subject }}\nEmail snippet:\n{{ $json.snippet }}\n\nYour response must be exactly one of these 8 options:\n1. To respond\n2. FYI\n3. Comment\n4. Notification\n5. Meeting update\n6. Awaiting reply\n7. Actioned\n8. Marketing",
        "batching": {},
        "promptType": "define"
      },
      "typeVersion": 1.7
    },
    {
      "id": "ddbdc31a-ea27-415e-b9e2-954959cfcb7a",
      "name": "收到新 Gmail 邮件",
      "type": "n8n-nodes-base.gmailTrigger",
      "position": [
        -560,
        340
      ],
      "parameters": {
        "filters": {},
        "pollTimes": {
          "item": [
            {
              "hour": 12
            }
          ]
        }
      },
      "credentials": {
        "gmailOAuth2": {
          "id": "V5NiRamelad0WTXx",
          "name": "Gmail account"
        }
      },
      "typeVersion": 1.2
    },
    {
      "id": "52b3a1d8-3be8-4f57-8759-7ae9abcdb5cb",
      "name": "获取可用 Gmail 标签",
      "type": "n8n-nodes-base.gmail",
      "position": [
        -140,
        220
      ],
      "webhookId": "8574ca24-b439-4bc3-aa28-ae2b539a8d8e",
      "parameters": {
        "resource": "label",
        "returnAll": true
      },
      "credentials": {
        "gmailOAuth2": {
          "id": "V5NiRamelad0WTXx",
          "name": "Gmail account"
        }
      },
      "typeVersion": 2.1
    },
    {
      "id": "2d15670f-5735-4110-a1eb-d5b6b8abef89",
      "name": "将标签名称映射至 ID",
      "type": "n8n-nodes-base.code",
      "position": [
        60,
        220
      ],
      "parameters": {
        "jsCode": "const items = $input.all();\nconst combinedObject = items.reduce((acc, item) => {\n  return { ...acc, [item.json.name]: item.json.id };\n}, {});\n\nreturn { combinedObject };\n"
      },
      "typeVersion": 2
    },
    {
      "id": "4c5d1e4a-93ce-4f78-b0ce-8b94aff99498",
      "name": "应用分类标签",
      "type": "n8n-nodes-base.gmail",
      "position": [
        960,
        320
      ],
      "webhookId": "2ffe45e7-a0e3-4304-a6bf-e217c15ea6cc",
      "parameters": {
        "labelIds": "={{ $('Map Label Names to IDs').item.json.combinedObject[$json.text] }}",
        "messageId": "={{ $('Merge Labels & Messages').item.json.id }}",
        "operation": "addLabels"
      },
      "credentials": {
        "gmailOAuth2": {
          "id": "V5NiRamelad0WTXx",
          "name": "Gmail account"
        }
      },
      "typeVersion": 2.1
    }
  ],
  "pinData": {},
  "connections": {
    "OpenRouter Chat Model": {
      "ai_languageModel": [
        [
          {
            "node": "Classify Email with AI",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "Classify Email with AI": {
      "main": [
        [
          {
            "node": "Apply Classification Label",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Map Label Names to IDs": {
      "main": [
        [
          {
            "node": "Merge Labels & Messages",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Merge Labels & Messages": {
      "main": [
        [
          {
            "node": "Classify Email with AI",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Aggregate Gmail Messages": {
      "main": [
        [
          {
            "node": "Fetch Available Gmail Labels",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "New Gmail Email Received": {
      "main": [
        [
          {
            "node": "Aggregate Gmail Messages",
            "type": "main",
            "index": 0
          },
          {
            "node": "Merge Labels & Messages",
            "type": "main",
            "index": 1
          }
        ]
      ]
    },
    "Fetch Available Gmail Labels": {
      "main": [
        [
          {
            "node": "Map Label Names to IDs",
            "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 - 其他, 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。

工作流信息
难度等级
中级
节点数量8
分类2
节点类型7
难度说明

适合有一定经验的用户,包含 6-15 个节点的中等复杂度工作流

作者
Alex Dunlop

Alex Dunlop

@alexdunlop

Kiwi Developer & Founder. Working on AI Startups while in London. Passionate about AI Tech, Learning, and Helping Engineers.

外部链接
在 n8n.io 查看

分享此工作流