InstaTest

工作流概述

这是一个包含11个节点的复杂工作流,主要用于自动化处理各种任务。

工作流源代码

下载
{
  "id": "qww129cm4TM9N8Ru",
  "meta": {
    "instanceId": "038da3428bba4563b42be267feeca21b4922693db254331ac640a5c56ee7cadf",
    "templateCredsSetupCompleted": true
  },
  "name": "InstaTest",
  "tags": [
    {
      "id": "8PlqXsDyqVlHJ7RC",
      "name": "AI",
      "createdAt": "2024-07-10T14:12:10.657Z",
      "updatedAt": "2024-07-10T14:12:10.657Z"
    }
  ],
  "nodes": [
    {
      "id": "51dcaa84-d1f9-4abc-aebc-24a06801e42d",
      "name": "Set your system promt for AI",
      "type": "n8n-nodes-base.set",
      "notes": "In this node in \"prompt\" variable you can set your system prompt",
      "position": [
        1120,
        620
      ],
      "parameters": {
        "options": {},
        "assignments": {
          "assignments": [
            {
              "id": "0b3c3d71-5627-4b8c-91f0-ac44eaedf196",
              "name": "prompt",
              "type": "string",
              "value": "=Persona: You are a instagram influencer.
Context: You receive a messages from your subscribers
Task: Answer questions in your writing style and patterns according to your previous posts text. Use your post only for style and patterns reference.
Style rules:
simple answers"
            },
            {
              "id": "c2a9e272-5c0d-4685-ad0e-ce6995f92a1c",
              "name": "sessionId",
              "type": "string",
              "value": "={{ $json.body.session_id }}"
            },
            {
              "id": "b3c20ee3-07a1-4584-b0d9-7310a2c6b723",
              "name": "chatInput",
              "type": "string",
              "value": "={{ $json.body.text }}"
            }
          ]
        }
      },
      "typeVersion": 3.3
    },
    {
      "id": "0fb36573-d632-4403-8809-3973f9caa32a",
      "name": "Local n8n memory",
      "type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
      "position": [
        1500,
        780
      ],
      "parameters": {
        "sessionKey": "={{ $('Set your system promt for AI').last().json.sessionId }}",
        "sessionIdType": "customKey",
        "contextWindowLength": 20
      },
      "typeVersion": 1.3
    },
    {
      "id": "2f0471a7-2a84-41ce-aab1-896d5ea95ac3",
      "name": "ChatGPT model",
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
      "position": [
        1360,
        780
      ],
      "parameters": {
        "options": {}
      },
      "credentials": {
        "openAiApi": {
          "id": "HxWZhtJcnqTXVHAA",
          "name": "General"
        }
      },
      "typeVersion": 1
    },
    {
      "id": "49abc3a3-faf9-4249-b874-908138a84aea",
      "name": "Send respond ",
      "type": "n8n-nodes-base.respondToWebhook",
      "position": [
        1720,
        620
      ],
      "parameters": {
        "options": {}
      },
      "typeVersion": 1.1
    },
    {
      "id": "49382508-9307-4ffa-8b31-78fac3a7db10",
      "name": "Sticky Note",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        320,
        360
      ],
      "parameters": {
        "color": 5,
        "width": 458.4028599661066,
        "height": 447.98321744507007,
        "content": "## Easy Instagram(via ManyChat) bot
---
### Description:
This template is a main part of Entire solution. It's getting new message from Instagram via ManyChat(Extra No-Code tool for getting and sending message in Instagram). Generating message using ChatGPT and send back to ManyChat that sends it to Instagrtam.

### Logic:
1. Getting message from Instagram(from ManyChat)
2. Set you system prompt for AI
3. Create simple answer for message in AI block
4. Send answer to Instagram(to ManyChat)

---
*Helpful links:*
- [Guide in Notion how to create full bot](https://shadowed-pound-d6e.notion.site/Instagram-GPT-light-version-Manychat-X-N8N-176293bddff880899a9ac255585d29f7?pvs=4)
- [ManyChat](https://manychat.partnerlinks.io/vm4wkw8j81tc)"
      },
      "typeVersion": 1
    },
    {
      "id": "5d14544c-7039-435f-a53c-615b5722bb99",
      "name": "Getting message from Instagram",
      "type": "n8n-nodes-base.webhook",
      "position": [
        900,
        620
      ],
      "webhookId": "68d3fbc9-6e49-4bdc-851c-2a532be911ab",
      "parameters": {
        "path": "instagram_chat",
        "options": {},
        "httpMethod": "POST",
        "responseMode": "responseNode"
      },
      "typeVersion": 2
    },
    {
      "id": "3770f558-341b-4d67-a7f0-0bb2fecf51a3",
      "name": "Sticky Note1",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1320,
        300
      ],
      "parameters": {
        "width": 313.9634922216307,
        "height": 614.7475040550845,
        "content": "## 3) AI block
---
There is 3 nodes:
- AI Agent
- Chat GPT model
- Memory for history messages

### To do:
- in ChatGPT node you can choose the best model for you
- in Memory Block you can change number of messages in history

"
      },
      "typeVersion": 1
    },
    {
      "id": "cbb6c5a2-9b96-4305-afce-5ac560ae2dec",
      "name": "AI Agent",
      "type": "@n8n/n8n-nodes-langchain.agent",
      "position": [
        1340,
        620
      ],
      "parameters": {
        "text": "={{ $json.chatInput }}",
        "options": {
          "systemMessage": "={{ $json.prompt }}"
        },
        "promptType": "define"
      },
      "typeVersion": 1.7
    },
    {
      "id": "4e28119f-b1aa-4b20-a8ed-28bd137f9627",
      "name": "Sticky Note2",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        820,
        360
      ],
      "parameters": {
        "height": 440,
        "content": "## 1) HTTP Post webhook

**To do:**
Just copy production link from this node and insert to custom action in ManyChat

No edits needed"
      },
      "typeVersion": 1
    },
    {
      "id": "b18a8890-b420-4086-91c8-8edbc845c8af",
      "name": "Sticky Note3",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1080,
        480
      ],
      "parameters": {
        "width": 220,
        "height": 320,
        "content": "## 2) Edit prompt

**To do:**
Go inside and change input
"
      },
      "typeVersion": 1
    },
    {
      "id": "74d4e6f5-069e-4b37-8005-8c03226b05df",
      "name": "Sticky Note4",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        1660,
        480
      ],
      "parameters": {
        "height": 300,
        "content": "## 4) Respond webhook

No edits needed"
      },
      "typeVersion": 1
    }
  ],
  "active": true,
  "pinData": {},
  "settings": {
    "executionOrder": "v1"
  },
  "versionId": "2f36fc7a-0a69-4af3-a958-25e9d278f058",
  "connections": {
    "AI Agent": {
      "main": [
        [
          {
            "node": "Send respond ",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "ChatGPT model": {
      "ai_languageModel": [
        [
          {
            "node": "AI Agent",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "Local n8n memory": {
      "ai_memory": [
        [
          {
            "node": "AI Agent",
            "type": "ai_memory",
            "index": 0
          }
        ]
      ]
    },
    "Set your system promt for AI": {
      "main": [
        [
          {
            "node": "AI Agent",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Getting message from Instagram": {
      "main": [
        [
          {
            "node": "Set your system promt for AI",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  }
}

功能特点

  • 自动检测新邮件
  • AI智能内容分析
  • 自定义分类规则
  • 批量处理能力
  • 详细的处理日志

技术分析

节点类型及作用

  • Set
  • @N8N/N8N Nodes Langchain.Memorybufferwindow
  • @N8N/N8N Nodes Langchain.Lmchatopenai
  • Respondtowebhook
  • Stickynote

复杂度评估

配置难度:
★★★★☆
维护难度:
★★☆☆☆
扩展性:
★★★★☆

实施指南

前置条件

  • 有效的Gmail账户
  • n8n平台访问权限
  • Google API凭证
  • AI分类服务订阅

配置步骤

  1. 在n8n中导入工作流JSON文件
  2. 配置Gmail节点的认证信息
  3. 设置AI分类器的API密钥
  4. 自定义分类规则和标签映射
  5. 测试工作流执行
  6. 配置定时触发器(可选)

关键参数

参数名称 默认值 说明
maxEmails 50 单次处理的最大邮件数量
confidenceThreshold 0.8 分类置信度阈值
autoLabel true 是否自动添加标签

最佳实践

优化建议

  • 定期更新AI分类模型以提高准确性
  • 根据邮件量调整处理批次大小
  • 设置合理的分类置信度阈值
  • 定期清理过期的分类规则

安全注意事项

  • 妥善保管API密钥和认证信息
  • 限制工作流的访问权限
  • 定期审查处理日志
  • 启用双因素认证保护Gmail账户

性能优化

  • 使用增量处理减少重复工作
  • 缓存频繁访问的数据
  • 并行处理多个邮件分类任务
  • 监控系统资源使用情况

故障排除

常见问题

邮件未被正确分类

检查AI分类器的置信度阈值设置,适当降低阈值或更新训练数据。

Gmail认证失败

确认Google API凭证有效且具有正确的权限范围,重新进行OAuth授权。

调试技巧

  • 启用详细日志记录查看每个步骤的执行情况
  • 使用测试邮件验证分类逻辑
  • 检查网络连接和API服务状态
  • 逐步执行工作流定位问题节点

错误处理

工作流包含以下错误处理机制:

  • 网络超时自动重试(最多3次)
  • API错误记录和告警
  • 处理失败邮件的隔离机制
  • 异常情况下的回滚操作