ChatGPT Visionを使用したスパムメールコンテンツの分析
上級
これはAI, SecOps分野の自動化ワークフローで、18個のノードを含みます。主にSet, Code, Jira, HttpRequest, GmailTriggerなどのノードを使用、AI技術を活用したスマート自動化を実現。 ChatGPT Vision を使用して疑わしいメールコンテンツを分析する
前提条件
- •ターゲットAPIの認証情報が必要な場合あり
- •Googleアカウント + Gmail API認証情報
- •OpenAI API Key
ワークフロープレビュー
ノード接続関係を可視化、ズームとパンをサポート
ワークフローをエクスポート
以下のJSON設定をn8nにインポートして、このワークフローを使用できます
{
"meta": {
"instanceId": "03e9d14e9196363fe7191ce21dc0bb17387a6e755dcc9acc4f5904752919dca8"
},
"nodes": [
{
"id": "1bad6bfc-9ec9-48a5-b8f7-73c4de3d08cf",
"name": "Gmail トリガー",
"type": "n8n-nodes-base.gmailTrigger",
"position": [
1480,
160
],
"parameters": {
"simple": false,
"filters": {},
"options": {},
"pollTimes": {
"item": [
{
"mode": "everyMinute"
}
]
}
},
"credentials": {
"gmailOAuth2": {
"id": "kkhNhqKpZt6IUZd0",
"name": " Gmail"
}
},
"typeVersion": 1.2
},
{
"id": "9ac747a1-4fd8-46ba-b4c1-75fd17aab2ed",
"name": "Microsoft Outlook Trigger",
"type": "n8n-nodes-base.microsoftOutlookTrigger",
"disabled": true,
"position": [
1480,
720
],
"parameters": {
"fields": [
"body",
"toRecipients",
"subject",
"bodyPreview"
],
"output": "fields",
"filters": {},
"options": {},
"pollTimes": {
"item": [
{
"mode": "everyMinute"
}
]
}
},
"credentials": {
"microsoftOutlookOAuth2Api": {
"id": "vTCK0oVQ0WjFrI5H",
"name": " Outlook Credential"
}
},
"typeVersion": 1
},
{
"id": "5bf9b0e8-b84e-44a2-aad2-45dde3e4ab1b",
"name": "Screenshot HTML",
"type": "n8n-nodes-base.httpRequest",
"position": [
2520,
480
],
"parameters": {
"url": "https://hcti.io/v1/image",
"method": "POST",
"options": {},
"sendBody": true,
"sendQuery": true,
"authentication": "genericCredentialType",
"bodyParameters": {
"parameters": [
{
"name": "html",
"value": "={{ $json.htmlBody }}"
}
]
},
"genericAuthType": "httpBasicAuth",
"queryParameters": {
"parameters": [
{}
]
}
},
"credentials": {
"httpBasicAuth": {
"id": "8tm8mUWmPvtmPFPk",
"name": "hcti.io"
}
},
"typeVersion": 4.2
},
{
"id": "fc770d1d-6c18-4d14-8344-1dc042464df6",
"name": "Retrieve Screenshot",
"type": "n8n-nodes-base.httpRequest",
"position": [
2700,
480
],
"parameters": {
"url": "={{ $json.url }}",
"options": {},
"authentication": "genericCredentialType",
"genericAuthType": "httpBasicAuth"
},
"credentials": {
"httpBasicAuth": {
"id": "8tm8mUWmPvtmPFPk",
"name": "hcti.io"
}
},
"typeVersion": 4.2
},
{
"id": "2f3e5cc0-24e8-450a-898b-71e2d6f7bb58",
"name": "設定 Outlook Variables",
"type": "n8n-nodes-base.set",
"position": [
2020,
720
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "38bd3db2-1a8d-4c40-a2dd-336e0cc84224",
"name": "htmlBody",
"type": "string",
"value": "={{ $('Microsoft Outlook Trigger').item.json.body.content }}"
},
{
"id": "13bdd95b-ef02-486e-b38b-d14bd05a4a8a",
"name": "headers",
"type": "string",
"value": "={{ $json}}"
},
{
"id": "20566ad4-7eb7-42b1-8a0d-f8b759610f10",
"name": "subject",
"type": "string",
"value": "={{ $('Microsoft Outlook Trigger').item.json.subject }}"
},
{
"id": "7171998f-a5a2-4e23-946a-9c1ad75710e7",
"name": "recipient",
"type": "string",
"value": "={{ $('Microsoft Outlook Trigger').item.json.toRecipients[0].emailAddress.address }}"
},
{
"id": "cc262634-2470-4524-8319-abe2518a6335",
"name": "textBody",
"type": "string",
"value": "={{ $('Retrieve Headers of Email').item.json.body.content }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "374e5b16-a666-4706-9fd2-762b2927012d",
"name": "設定 Gmail Variables",
"type": "n8n-nodes-base.set",
"position": [
2040,
160
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "38bd3db2-1a8d-4c40-a2dd-336e0cc84224",
"name": "htmlBody",
"type": "string",
"value": "={{ $json.html }}"
},
{
"id": "18fbcf78-6d3c-4036-b3a2-fb5adf22176a",
"name": "headers",
"type": "string",
"value": "={{ $json.headers }}"
},
{
"id": "1d690098-be2a-4604-baf8-62f314930929",
"name": "subject",
"type": "string",
"value": "={{ $json.subject }}"
},
{
"id": "8009f00a-547f-4eb1-b52d-2e7305248885",
"name": "recipient",
"type": "string",
"value": "={{ $json.to.text }}"
},
{
"id": "1932e97d-b03b-4964-b8bc-8262aaaa1f7a",
"name": "textBody",
"type": "string",
"value": "={{ $json.text }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "3166738e-d0a3-475b-8b19-51afd519ee3a",
"name": "Retrieve Headers of メール",
"type": "n8n-nodes-base.httpRequest",
"position": [
1680,
720
],
"parameters": {
"url": "=https://graph.microsoft.com/v1.0/me/messages/{{ $json.id }}?$select=internetMessageHeaders,body",
"options": {},
"sendHeaders": true,
"authentication": "predefinedCredentialType",
"headerParameters": {
"parameters": [
{
"name": "Accept",
"value": "application/json"
},
{
"name": "Prefer",
"value": "outlook.body-content-type=\"text\""
}
]
},
"nodeCredentialType": "microsoftOutlookOAuth2Api"
},
"credentials": {
"microsoftOutlookOAuth2Api": {
"id": "vTCK0oVQ0WjFrI5H",
"name": " Outlook Credential"
}
},
"typeVersion": 4.2
},
{
"id": "25ae222c-088f-4565-98d6-803c8c1b0826",
"name": "Format Headers",
"type": "n8n-nodes-base.code",
"position": [
1860,
720
],
"parameters": {
"jsCode": "const input = $('Retrieve Headers of Email').item.json.internetMessageHeaders;\n\nconst result = input.reduce((acc, { name, value }) => {\n if (!acc[name]) acc[name] = [];\n acc[name].push(value);\n return acc;\n}, {});\n\nreturn result;"
},
"typeVersion": 2
},
{
"id": "8f14f267-1074-43ea-968d-26a6ab36fd7b",
"name": "設定 Email Variables",
"type": "n8n-nodes-base.set",
"position": [
2360,
480
],
"parameters": {
"options": {},
"includeOtherFields": true
},
"typeVersion": 3.4
},
{
"id": "45d156aa-91f4-483c-91d4-c9de4a4f595d",
"name": "ChatGPT Analysis",
"type": "@n8n/n8n-nodes-langchain.openAi",
"position": [
3100,
480
],
"parameters": {
"text": "=Describe this image. Determine if the email could be a phishing email. The message headers are as follows:\n{{ $('Set Email Variables').item.json.headers }}\n\nFormat the response for Jira who uses a wiki-style renderer. Do not include ``` around your response.",
"modelId": {
"__rl": true,
"mode": "list",
"value": "chatgpt-4o-latest",
"cachedResultName": "CHATGPT-4O-LATEST"
},
"options": {
"maxTokens": 1500
},
"resource": "image",
"inputType": "base64",
"operation": "analyze"
},
"credentials": {
"openAiApi": {
"id": "76",
"name": "OpenAi account"
}
},
"typeVersion": 1.6
},
{
"id": "62ca591b-6627-496c-96a7-95cb0081480d",
"name": "Create Jira Ticket",
"type": "n8n-nodes-base.jira",
"position": [
3500,
480
],
"parameters": {
"project": {
"__rl": true,
"mode": "list",
"value": "10001",
"cachedResultName": "Support"
},
"summary": "=Phishing Email Reported: \"{{ $('Set Email Variables').item.json.subject }}\"",
"issueType": {
"__rl": true,
"mode": "list",
"value": "10008",
"cachedResultName": "Task"
},
"additionalFields": {
"description": "=A phishing email was reported by {{ $('Set Email Variables').item.json.recipient }} with the subject line \"{{ $('Set Email Variables').item.json.subject }}\" and body:\n{{ $('Set Email Variables').item.json.textBody }}\n\\\\\n\\\\\n\\\\\nh2. Here is ChatGPT's analysis of the email:\n{{ $json.content }}"
}
},
"credentials": {
"jiraSoftwareCloudApi": {
"id": "BZmmGUrNIsgM9fDj",
"name": "New Jira Cloud"
}
},
"typeVersion": 1
},
{
"id": "071380c8-8070-4f8f-86c6-87c4ee3bc261",
"name": "Rename Screenshot",
"type": "n8n-nodes-base.code",
"position": [
3680,
480
],
"parameters": {
"mode": "runOnceForEachItem",
"jsCode": "$('Retrieve Screenshot').item.binary.data.fileName = 'emailScreenshot.png'\n\nreturn $('Retrieve Screenshot').item;"
},
"typeVersion": 2
},
{
"id": "05c57490-c1ee-48f0-9e38-244c9a995e22",
"name": "Upload Screenshot of メール to Jira",
"type": "n8n-nodes-base.jira",
"position": [
3860,
480
],
"parameters": {
"issueKey": "={{ $('Create Jira Ticket').item.json.key }}",
"resource": "issueAttachment"
},
"credentials": {
"jiraSoftwareCloudApi": {
"id": "BZmmGUrNIsgM9fDj",
"name": "New Jira Cloud"
}
},
"typeVersion": 1
},
{
"id": "be02770d-a943-41f5-98a9-5c433a6a3dbf",
"name": "付箋",
"type": "n8n-nodes-base.stickyNote",
"position": [
1420,
-107.36679523834897
],
"parameters": {
"color": 7,
"width": 792.3026315789474,
"height": 426.314163659402,
"content": "\n## Gmail Integration and Data Extraction\n\nThis section of the workflow connects to a Gmail account using the **Gmail Trigger** node, capturing incoming emails in real-time, with checks performed every minute. Once an email is detected, its key components—such as the subject, recipient, body, and headers—are extracted and assigned to variables using the **Set Gmail Variables** node. These variables are structured for subsequent analysis and processing in later steps."
},
"typeVersion": 1
},
{
"id": "c1d2f691-669a-46de-9ef8-59ce4e6980c5",
"name": "付箋1",
"type": "n8n-nodes-base.stickyNote",
"position": [
1420,
380.6918768014301
],
"parameters": {
"color": 7,
"width": 792.3026315789474,
"height": 532.3344389880435,
"content": "\n## Microsoft Outlook Integration and Email Header Processing\n\nThis section connects to a Microsoft Outlook account to monitor incoming emails using the **Microsoft Outlook Trigger** node, which checks for new messages every minute. Emails are then processed to retrieve detailed headers and body content via the **Retrieve Headers of Email** node. The headers are structured into a user-friendly format using the **Format Headers** code node, ensuring clarity for further analysis. Key details, including the email's subject, recipient, and body content, are assigned to variables with the **Set Outlook Variables** node for streamlined integration into subsequent workflow steps."
},
"typeVersion": 1
},
{
"id": "c189e2e0-9f51-4bc0-a483-8b7f0528be70",
"name": "付箋2",
"type": "n8n-nodes-base.stickyNote",
"position": [
2287.3684210526317,
46.18421052631584
],
"parameters": {
"color": 7,
"width": 580.4605263157906,
"height": 615.460526315789,
"content": "\n## HTML Screenshot Generation and Email Visualization\n\nThis section processes an email’s HTML content to create a visual representation, useful for documentation or phishing detection workflows. The **Set Email Variables** node organizes the email's HTML body into a format ready for processing. The **Screenshot HTML** node sends this HTML content to the **hcti.io** API, which generates a screenshot of the email's layout. The **Retrieve Screenshot** node then fetches the image URL for further use in the workflow. This setup ensures that the email's appearance is preserved in a visually accessible format, simplifying review and reporting. Keep in mind however that this exposes the email content to a third party. If you self host n8n, you can deploy a cli tool to rasterize locally instead."
},
"typeVersion": 1
},
{
"id": "9076f9e9-f4fb-409a-9580-1ae459094c31",
"name": "付箋3",
"type": "n8n-nodes-base.stickyNote",
"position": [
2880,
123.72476075009968
],
"parameters": {
"color": 7,
"width": 507.82894736842223,
"height": 537.9199760920052,
"content": "\n## AI-Powered Email Analysis with ChatGPT\n\nThis section leverages AI to analyze email content and headers for phishing indicators. The **ChatGPT Analysis** node utilizes the ChatGPT-4 model to review the email screenshot and associated metadata, including message headers. It generates a detailed report indicating whether the email might be a phishing attempt. The output is formatted specifically for Jira’s wiki-style renderer, making it ready for seamless integration into ticketing workflows. This ensures thorough and automated email threat assessments."
},
"typeVersion": 1
},
{
"id": "ca2488af-e787-4675-802a-8b4f2d845376",
"name": "付箋4",
"type": "n8n-nodes-base.stickyNote",
"position": [
3400,
122.88662032580646
],
"parameters": {
"color": 7,
"width": 692.434210526317,
"height": 529.5475902005091,
"content": "\n## Automated Jira Ticket Creation for Phishing Reports\n\nThis section streamlines the process of reporting phishing emails by automatically creating detailed Jira tickets. The **Create Jira Ticket** node compiles email information, including the subject, recipient, body text, and ChatGPT's phishing analysis, into a structured ticket. The **Rename Screenshot** node ensures that the email screenshot file is appropriately labeled for attachment. Finally, the **Upload Screenshot of Email to Jira** node attaches the email’s visual representation to the ticket, providing additional context for the security team. This integration ensures that phishing reports are logged with all necessary details, enabling efficient tracking and resolution."
},
"typeVersion": 1
}
],
"pinData": {},
"connections": {
"Gmail Trigger": {
"main": [
[
{
"node": "Set Gmail Variables",
"type": "main",
"index": 0
}
]
]
},
"25ae222c-088f-4565-98d6-803c8c1b0826": {
"main": [
[
{
"node": "Set Outlook Variables",
"type": "main",
"index": 0
}
]
]
},
"5bf9b0e8-b84e-44a2-aad2-45dde3e4ab1b": {
"main": [
[
{
"node": "fc770d1d-6c18-4d14-8344-1dc042464df6",
"type": "main",
"index": 0
}
]
]
},
"45d156aa-91f4-483c-91d4-c9de4a4f595d": {
"main": [
[
{
"node": "62ca591b-6627-496c-96a7-95cb0081480d",
"type": "main",
"index": 0
}
]
]
},
"071380c8-8070-4f8f-86c6-87c4ee3bc261": {
"main": [
[
{
"node": "Upload Screenshot of Email to Jira",
"type": "main",
"index": 0
}
]
]
},
"62ca591b-6627-496c-96a7-95cb0081480d": {
"main": [
[
{
"node": "071380c8-8070-4f8f-86c6-87c4ee3bc261",
"type": "main",
"index": 0
}
]
]
},
"fc770d1d-6c18-4d14-8344-1dc042464df6": {
"main": [
[
{
"node": "45d156aa-91f4-483c-91d4-c9de4a4f595d",
"type": "main",
"index": 0
}
]
]
},
"Set Email Variables": {
"main": [
[
{
"node": "5bf9b0e8-b84e-44a2-aad2-45dde3e4ab1b",
"type": "main",
"index": 0
}
]
]
},
"Set Gmail Variables": {
"main": [
[
{
"node": "Set Email Variables",
"type": "main",
"index": 0
}
]
]
},
"Set Outlook Variables": {
"main": [
[
{
"node": "Set Email Variables",
"type": "main",
"index": 0
}
]
]
},
"9ac747a1-4fd8-46ba-b4c1-75fd17aab2ed": {
"main": [
[
{
"node": "Retrieve Headers of Email",
"type": "main",
"index": 0
}
]
]
},
"Retrieve Headers of Email": {
"main": [
[
{
"node": "25ae222c-088f-4565-98d6-803c8c1b0826",
"type": "main",
"index": 0
}
]
]
}
}
}よくある質問
このワークフローの使い方は？
上記のJSON設定コードをコピーし、n8nインスタンスで新しいワークフローを作成して「JSONからインポート」を選択、設定を貼り付けて認証情報を必要に応じて変更してください。
このワークフローはどんな場面に適していますか？
上級 - 人工知能, セキュリティ運用
有料ですか？
このワークフローは完全無料です。ただし、ワークフローで使用するサードパーティサービス（OpenAI APIなど）は別途料金が発生する場合があります。
関連ワークフロー
ChatGPTを使って suspicious メールコンテンツを分析し分類する
可疑メール本文をChatGPTで分析して分類する
If
Set
Code
+
If
Set
Code
25 ノードAngel Menendez
人工知能
Gmailのメールヘッダーを解析してIP信頼性とスプーフィングを検出
Gmailのメールヘッダーを解析し、IP評判とスプーフィングを検出する
If
Set
Code
+
If
Set
Code
40 ノードAngel Menendez
セキュリティ運用
Outlookのメールヘッダーを解析してIP信頼性とスプーフィングを検出
Outlookのメールヘッダーを解析し、IP評判とスプーフィングを検出する
If
Set
Code
+
If
Set
Code
41 ノードAngel Menendez
セキュリティ運用
Qualys スラックサイニスセーピボットによるセキュリティ運用の強化
Qualys Slackワイヤーロボを使ってセキュリティ運用の効率を向上
Set
Switch
Webhook
+
Set
Switch
Webhook
23 ノードAngel Menendez
人工知能
CallForge - 03 - Gong転写プロセッサーとSalesforceの強化
CallForge - 03 - Gongの文字起こしプロセッサーとSalesforceのエンハンスメント
Set
Code
Merge
+
Set
Code
Merge
23 ノードAngel Menendez
営業
HRMateの修理
LlamaIndexとGPT-4o-miniを使用して、応募者筛选を自動化し、パーソナライズされたメールによる返信を生成
If
Set
Code
+
If
Set
Code
30 ノードKhairul Muhtadin
人事
ワークフロー情報
難易度
上級
ノード数18
カテゴリー2
ノードタイプ8
作成者
Angel Menendez
@djangelicAngel Menendez is a Staff Developer Advocate at n8n.io, specializing in low-code tools for cybersecurity workflows. From Puerto Rico, Angel's tech journey began by helping his father translate technical books. He later started a web development business and transitioned from a career as a flight attendant to cybersecurity engineering. His workflows have saved companies significant time. Outside work, Angel enjoys time with his two sons, riding electric bikes, reading, and exploring new places.
外部リンク
n8n.ioで表示 →
このワークフローを共有