音声サポートワークフロー
上級
これはOther, AI分野の自動化ワークフローで、22個のノードを含みます。主にIf, Set, Code, Switch, Telegram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。 AI音声コーチとジェネレーター：Telegram、OpenAI、Geminiを活用
前提条件
- •Telegram Bot Token
- •OpenAI API Key
- •Google Gemini API Key
使用ノード (22)
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
{
"id": "Fdbft9uw8mLGXMoE",
"meta": {
"instanceId": "13d96e1ebd7901d1ed300d36db3a4447107e9ad60df51fe711e45683875362aa",
"templateCredsSetupCompleted": true
},
"name": "Speech Support Workflow",
"tags": [
{
"id": "88Rkm7VaAFefsT34",
"name": "AI",
"createdAt": "2025-05-06T22:52:26.053Z",
"updatedAt": "2025-05-06T22:52:26.053Z"
},
{
"id": "s1UA6FThbKhQYbLu",
"name": "MultiModal",
"createdAt": "2025-05-06T22:52:35.914Z",
"updatedAt": "2025-05-06T22:52:35.914Z"
},
{
"id": "ANT04PP2WxQmkjzl",
"name": "Integrations",
"createdAt": "2025-05-06T22:53:02.798Z",
"updatedAt": "2025-05-06T22:53:02.798Z"
}
],
"nodes": [
{
"id": "8868fc75-4a21-4900-b2b9-7860ee981a9e",
"name": "AIエージェント",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
1640,
240
],
"parameters": {
"text": "={{ $('Route Flow Based on Message Content').item.json.text }}",
"options": {
"systemMessage": "={{ $json.system_prompt }}\n\nYou are generating text for a Telegram message. The text should be plain. No * or **"
},
"promptType": "define"
},
"typeVersion": 1.8
},
{
"id": "23f48680-a190-48a5-bb7c-e070db41b9e7",
"name": "Google Gemini チャットモデル",
"type": "@n8n/n8n-nodes-langchain.lmChatGoogleGemini",
"position": [
1620,
800
],
"parameters": {
"options": {},
"modelName": "models/gemini-2.0-flash-001"
},
"credentials": {
"googlePalmApi": {
"id": "zCkkU4GKPR7wANF5",
"name": "Google Gemini(PaLM) Api account"
}
},
"typeVersion": 1
},
{
"id": "6a909fb0-f550-4b5e-94db-6e16682d70bd",
"name": "Telegramメッセージ受信",
"type": "n8n-nodes-base.telegramTrigger",
"position": [
-480,
240
],
"webhookId": "20140af0-c902-44db-9c53-051def981f9a",
"parameters": {
"updates": [
"message"
],
"additionalFields": {}
},
"credentials": {
"telegramApi": {
"id": "WvBkWguhZJQm5FpM",
"name": "Telegram account"
}
},
"typeVersion": 1.2
},
{
"id": "73d19e09-efc4-43c4-a4e9-382ae66c7651",
"name": "テキストまたは音声メッセージの確認",
"type": "n8n-nodes-base.set",
"position": [
-260,
240
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "b37e51e5-e2c7-4328-b02b-80d08164d595",
"name": "text",
"type": "string",
"value": "={{ $json.message.text||\"\" }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "a7ade841-258d-45b2-9150-a56490a4c37f",
"name": "音声ファイルのダウンロード",
"type": "n8n-nodes-base.telegram",
"position": [
180,
120
],
"webhookId": "68e0f93e-5dd0-41aa-89e4-4e7a6be9d3b2",
"parameters": {
"fileId": "={{ $('Recieve Telegram Message').item.json.message.voice.file_id }}",
"resource": "file"
},
"credentials": {
"telegramApi": {
"id": "WvBkWguhZJQm5FpM",
"name": "Telegram account"
}
},
"typeVersion": 1.2
},
{
"id": "73cb448e-f00e-4879-8fa2-facb259b76b2",
"name": "音声ファイルの文字起こし",
"type": "@n8n/n8n-nodes-langchain.openAi",
"position": [
400,
120
],
"parameters": {
"options": {},
"resource": "audio",
"operation": "transcribe"
},
"credentials": {
"openAiApi": {
"id": "cDXozPn1syyex1aJ",
"name": "OpenAi account"
}
},
"typeVersion": 1.8
},
{
"id": "a1999ecd-cabf-4740-a9a6-98486a868b7f",
"name": "音声メッセージの場合",
"type": "n8n-nodes-base.if",
"position": [
-60,
240
],
"parameters": {
"options": {},
"conditions": {
"options": {
"version": 2,
"leftValue": "",
"caseSensitive": true,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "d60f6ce2-afd0-4ee1-a7c3-3d5bbdb68ea2",
"operator": {
"type": "string",
"operation": "empty",
"singleValue": true
},
"leftValue": "={{ $json.text }}",
"rightValue": ""
}
]
}
},
"typeVersion": 2.2
},
{
"id": "ac5cad34-3756-4fe3-9269-ae96e5b49e8f",
"name": "LLM応答から不要な文字を除去するコード",
"type": "n8n-nodes-base.code",
"position": [
2060,
240
],
"parameters": {
"language": "python",
"pythonCode": "import re\n\ndef clean_markdown_for_telegram(text):\n \"\"\"\n Removes common Markdown formatting characters from a string.\n\n Args:\n text: The input string.\n\n Returns:\n A new string with Markdown characters removed.\n \"\"\"\n markdown_chars = r\"[*_~`\\[\\]()#+\\-=|{}.!]\"\n cleaned_text = re.sub(markdown_chars, \"\", text)\n cleaned_text = \" \".join(cleaned_text.split()).strip()\n return cleaned_text\n\n# Loop over input items and create new items with the cleaned text\noutput_items = []\nfor item in _input.all():\n feedback_text = item.json.get(\"output\", \"\")\n cleaned_feedback = clean_markdown_for_telegram(feedback_text)\n output_items.append({\"json\": {\"cleanedText\": cleaned_feedback}})\n\nreturn output_items"
},
"typeVersion": 2
},
{
"id": "82761634-7472-4ce1-806a-2b80aca985e3",
"name": "出力を4000文字以下のチャンクに分割するコード",
"type": "n8n-nodes-base.code",
"position": [
2280,
240
],
"parameters": {
"language": "python",
"pythonCode": "def split_text_for_telegram(text, max_length=4000):\n \"\"\"\n Splits a long text into a list of strings, each with a maximum length\n suitable for Telegram messages.\n\n Args:\n text: The input string to split.\n max_length: The maximum length of each resulting string (default: 4000).\n\n Returns:\n A list of strings, where each string is a chunk of the original text\n with a maximum length of max_length.\n \"\"\"\n if len(text) <= max_length:\n return [text]\n\n chunks = []\n start_index = 0\n while start_index < len(text):\n end_index = min(start_index + max_length, len(text))\n\n split_point = end_index\n if end_index < len(text):\n last_sentence_end = -1\n for i in range(start_index + max_length - 1, start_index - 1, -1):\n if i < len(text) and text[i] in ['.', '?', '!']:\n last_sentence_end = i + 1\n break\n if last_sentence_end > start_index:\n split_point = last_sentence_end\n\n chunks.append(text[start_index:split_point])\n start_index = split_point\n\n return chunks\n\noutput_items = []\nmax_length = 4000\n\nfor item in _input.all():\n text = item.json.get(\"cleanedText\", \"\")\n text_chunks = split_text_for_telegram(text, max_length)\n for chunk in text_chunks:\n output_items.append({\"json\": {\"telegramTextChunk\": chunk}})\n\nreturn output_items"
},
"typeVersion": 2
},
{
"id": "9adbfd4c-bbb9-4c92-bf07-a3b50a92aa02",
"name": "Telegramメッセージへ返信",
"type": "n8n-nodes-base.telegram",
"position": [
2500,
240
],
"webhookId": "4c77b108-e066-4538-986a-7535143cfaac",
"parameters": {
"text": "={{ $json.telegramTextChunk }}",
"chatId": "={{ $('Recieve Telegram Message').item.json.message.chat.id }}",
"additionalFields": {
"appendAttribution": false
}
},
"credentials": {
"telegramApi": {
"id": "WvBkWguhZJQm5FpM",
"name": "Telegram account"
}
},
"typeVersion": 1.2
},
{
"id": "5bdb6ec5-339e-4e8d-a746-9cdbe4d5f12f",
"name": "会話メモリの消去",
"type": "@n8n/n8n-nodes-langchain.memoryManager",
"position": [
940,
-20
],
"parameters": {
"mode": "delete",
"deleteMode": "all"
},
"typeVersion": 1.1
},
{
"id": "aae703f4-e891-4681-aae4-c426ebba5146",
"name": "会話メモリの保存",
"type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
"position": [
1120,
800
],
"parameters": {
"sessionKey": "={{ $('Recieve Telegram Message').item.json.message.from.id }}",
"sessionIdType": "customKey",
"contextWindowLength": 25
},
"typeVersion": 1.3
},
{
"id": "7457f085-9b19-4a00-9ad6-af2ca8ee16d5",
"name": "新規スピーチ開始用プロンプト設定",
"type": "n8n-nodes-base.set",
"position": [
1340,
-20
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "d5b33b03-241b-4193-915a-4eb4dfef05e9",
"name": "system_prompt",
"type": "string",
"value": "\"I am preparing to give a speech. Your role is to act as my speech preparation assistant. Please guide me through the process of getting ready to deliver this speech effectively. Ask me relevant questions and suggest steps we should take to ensure a successful presentation.\n\nPotential areas we can work on include:\n\nDefining the core message and key takeaways.\nUnderstanding the audience's needs and expectations.\nStructuring the speech for maximum impact.\nCrafting engaging content and supporting materials.\nDeveloping effective opening and closing remarks.\nPracticing delivery and managing speaking anxiety.\nAnticipating potential questions from the audience.\nConsidering the logistics of the presentation (e.g., time limits, equipment).\nWhere should we begin?"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "0a9f2c04-7f7e-488c-866c-717a78bf7db1",
"name": "フィードバックに基づくスピーチ生成用プロンプト設定",
"type": "n8n-nodes-base.set",
"position": [
1340,
220
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "ddbff433-f4bb-4ac1-a954-7fb32c942a9b",
"name": "system_prompt",
"type": "string",
"value": "I want you to act as a speech synthesizer and improvement agent. You have access to the content of several speeches I have previously provided, along with the constructive feedback I received on each and with this information your task is to generate a new speech.\n\nThis new speech should incorporate the following:\n\nKey themes and ideas that were present and well-received in my previous speeches.\nStructural elements and transitions that were identified as effective in past feedback.\nEngagement techniques that were noted as successful.\nAvoidance of areas for improvement highlighted in the feedback (e.g., rambling sections, unclear points, pacing issues).\nIncorporation of specific suggestions for improvement that were given.\nA similar tone and style to my previous speeches, while aiming for enhanced clarity and impact based on the feedback.\nPlease provide the complete text of the new speech. Feel free to ask clarifying questions if needed about the new topic, audience, or goal, or if you need a reminder of specific feedback points from my previous speeches. I am ready when you are.\""
}
]
}
},
"typeVersion": 3.4
},
{
"id": "f242361e-6bf4-4c4e-8cd0-72da06823842",
"name": "スピーチへのフィードバック提供用プロンプト設定",
"type": "n8n-nodes-base.set",
"position": [
1340,
420
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "b86d5905-872a-40fa-9855-054cd8991a0d",
"name": "system_prompt",
"type": "string",
"value": "I'd like you to act as a speech feedback agent. I will deliver a speech to you, and I want you to provide constructive criticism and insights on various aspects of my delivery and content. Please pay attention to: Clarity and Conciseness: Was the message easy to understand? Were there any parts that felt rambling or unnecessary? Engagement: How engaging was the speech overall? Were there moments where your attention might have drifted? Structure and Flow: Did the speech progress logically? Were the transitions smooth? Pacing and Timing: Was the speech delivered at an appropriate pace? Did it feel rushed or too slow? Vocal Delivery (if applicable): (If you are able to describe vocal elements) How was the tone, pitch, and volume? Did it enhance or detract from the message? Content and Impact: Was the content compelling and relevant? Did the speech achieve its intended purpose (as I will describe beforehand)? What was the overall impact of the message? Strengths: What were the most effective aspects of the speech? Areas for Improvement: What specific suggestions do you have to make the speech even better? Before I begin, I will briefly tell you the topic, my intended audience, and my goal for the speech. Are you ready?"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "c1f69bc3-2c93-404e-b338-82663deb975b",
"name": "メッセージ内容に基づくフロー振り分け",
"type": "n8n-nodes-base.switch",
"position": [
680,
260
],
"parameters": {
"rules": {
"values": [
{
"outputKey": "new_speech",
"conditions": {
"options": {
"version": 2,
"leftValue": "",
"caseSensitive": false,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "8b36fb19-1a5a-4fe1-aec2-7de8b5829972",
"operator": {
"type": "string",
"operation": "contains"
},
"leftValue": "={{ $json.text }}",
"rightValue": "new speech"
}
]
},
"renameOutput": true
},
{
"outputKey": "generate_speech",
"conditions": {
"options": {
"version": 2,
"leftValue": "",
"caseSensitive": false,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "708d114d-1146-4d8a-b972-cfb5a53a8d77",
"operator": {
"name": "filter.operator.equals",
"type": "string",
"operation": "equals"
},
"leftValue": "={{ $json.text }}",
"rightValue": "generate speech"
}
]
},
"renameOutput": true
}
]
},
"options": {
"ignoreCase": true,
"fallbackOutput": "extra"
}
},
"typeVersion": 3.2
},
{
"id": "6cf41698-2345-4ff8-bd1e-9549e372b454",
"name": "付箋",
"type": "n8n-nodes-base.stickyNote",
"position": [
1280,
-300
],
"parameters": {
"width": 220,
"height": 900,
"content": "## Dynamic System Prompting:\n\nThis node sets the AI's system prompt according to the user's request identified in the incoming message."
},
"typeVersion": 1
},
{
"id": "1d967631-9300-4fb6-b488-8c09beccbb05",
"name": "付箋1",
"type": "n8n-nodes-base.stickyNote",
"position": [
2000,
-300
],
"parameters": {
"width": 440,
"height": 720,
"content": "## Telegram-Ready Output: Formatting and Length Management:\n\nThese code nodes perform two crucial tasks:\n1. **Formatting:** Removing characters that could cause issues with Telegram's message parsing.\n2. **Chunking:** Dividing messages longer than Telegram's 4000-character limit into multiple shorter messages for sequential delivery."
},
"typeVersion": 1
},
{
"id": "c480550b-c94f-40ee-8338-3deb6bea28d8",
"name": "付箋2",
"type": "n8n-nodes-base.stickyNote",
"position": [
900,
-300
],
"parameters": {
"width": 340,
"height": 420,
"content": "## Clearing AI Agent Memory:\n\nThis node clears the AI agent's short-term memory. This helps to minimize the influence of past interactions on future responses, thereby reducing the likelihood of the AI generating inaccurate or irrelevant information (hallucinations)."
},
"typeVersion": 1
},
{
"id": "bd36f9a8-30d9-4e0a-978b-a74806685adc",
"name": "付箋3",
"type": "n8n-nodes-base.stickyNote",
"position": [
-320,
-300
],
"parameters": {
"width": 1160,
"height": 740,
"content": "## Processing Telegram Messages:\n\nThis section handles incoming messages from Telegram. It first checks if the message contains text.\n\n1. **Text Message:** If the message includes text, it's directly routed to the analysis switch node.\n2. **Audio Message:** If the message is an audio file:\n * The audio file is downloaded.\n * The audio is transcribed into text.\n * The transcribed text is then sent to the analysis switch node.\n\nFinally, the analyzed text (whether directly from a text message or transcribed from audio) is forwarded for further processing based on the analysis results."
},
"typeVersion": 1
},
{
"id": "cf7d8897-b963-4bb5-9b0e-3ab628e478c7",
"name": "付箋4",
"type": "n8n-nodes-base.stickyNote",
"position": [
1620,
-300
],
"parameters": {
"width": 300,
"height": 740,
"content": "## Gemini-Powered Response and Conversation Storage:\n\nThis node utilizes the Google Gemini model to generate a response to the user's prompt and stores the ongoing conversation."
},
"typeVersion": 1
},
{
"id": "9fffb613-536e-4b4d-8841-7b4bd3121eab",
"name": "付箋5",
"type": "n8n-nodes-base.stickyNote",
"position": [
-300,
-500
],
"parameters": {
"color": 4,
"width": 2740,
"height": 140,
"content": "## This n8n workflow acts as your personal AI speechwriting coach, directly accessible through Telegram. It listens to your spoken or typed drafts, provides insightful feedback on clarity, engagement, structure, and content, and iteratively refines your message based on your updates. Once you're ready, it synthesizes a brand-new speech or talk incorporating all the improvements and your accumulated ideas. This tool streamlines the speechwriting process, offering on-demand AI assistance to help you craft impactful and well-structured presentations."
},
"typeVersion": 1
}
],
"active": false,
"pinData": {},
"settings": {
"executionOrder": "v1"
},
"versionId": "4aaa8457-2661-4261-a601-0a0ffaffacff",
"connections": {
"8868fc75-4a21-4900-b2b9-7860ee981a9e": {
"main": [
[
{
"node": "ac5cad34-3756-4fe3-9269-ae96e5b49e8f",
"type": "main",
"index": 0
}
]
]
},
"a1999ecd-cabf-4740-a9a6-98486a868b7f": {
"main": [
[
{
"node": "a7ade841-258d-45b2-9150-a56490a4c37f",
"type": "main",
"index": 0
}
],
[
{
"node": "c1f69bc3-2c93-404e-b338-82663deb975b",
"type": "main",
"index": 0
}
]
]
},
"a7ade841-258d-45b2-9150-a56490a4c37f": {
"main": [
[
{
"node": "73cb448e-f00e-4879-8fa2-facb259b76b2",
"type": "main",
"index": 0
}
]
]
},
"73cb448e-f00e-4879-8fa2-facb259b76b2": {
"main": [
[
{
"node": "c1f69bc3-2c93-404e-b338-82663deb975b",
"type": "main",
"index": 0
}
]
]
},
"23f48680-a190-48a5-bb7c-e070db41b9e7": {
"ai_languageModel": [
[
{
"node": "8868fc75-4a21-4900-b2b9-7860ee981a9e",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"6a909fb0-f550-4b5e-94db-6e16682d70bd": {
"main": [
[
{
"node": "73d19e09-efc4-43c4-a4e9-382ae66c7651",
"type": "main",
"index": 0
}
]
]
},
"5bdb6ec5-339e-4e8d-a746-9cdbe4d5f12f": {
"main": [
[
{
"node": "7457f085-9b19-4a00-9ad6-af2ca8ee16d5",
"type": "main",
"index": 0
}
]
]
},
"aae703f4-e891-4681-aae4-c426ebba5146": {
"ai_memory": [
[
{
"node": "8868fc75-4a21-4900-b2b9-7860ee981a9e",
"type": "ai_memory",
"index": 0
},
{
"node": "5bdb6ec5-339e-4e8d-a746-9cdbe4d5f12f",
"type": "ai_memory",
"index": 0
}
]
]
},
"73d19e09-efc4-43c4-a4e9-382ae66c7651": {
"main": [
[
{
"node": "a1999ecd-cabf-4740-a9a6-98486a868b7f",
"type": "main",
"index": 0
}
]
]
},
"7457f085-9b19-4a00-9ad6-af2ca8ee16d5": {
"main": [
[
{
"node": "8868fc75-4a21-4900-b2b9-7860ee981a9e",
"type": "main",
"index": 0
}
]
]
},
"c1f69bc3-2c93-404e-b338-82663deb975b": {
"main": [
[
{
"node": "5bdb6ec5-339e-4e8d-a746-9cdbe4d5f12f",
"type": "main",
"index": 0
}
],
[
{
"node": "0a9f2c04-7f7e-488c-866c-717a78bf7db1",
"type": "main",
"index": 0
}
],
[
{
"node": "f242361e-6bf4-4c4e-8cd0-72da06823842",
"type": "main",
"index": 0
}
]
]
},
"f242361e-6bf4-4c4e-8cd0-72da06823842": {
"main": [
[
{
"node": "8868fc75-4a21-4900-b2b9-7860ee981a9e",
"type": "main",
"index": 0
}
]
]
},
"ac5cad34-3756-4fe3-9269-ae96e5b49e8f": {
"main": [
[
{
"node": "82761634-7472-4ce1-806a-2b80aca985e3",
"type": "main",
"index": 0
}
]
]
},
"0a9f2c04-7f7e-488c-866c-717a78bf7db1": {
"main": [
[
{
"node": "8868fc75-4a21-4900-b2b9-7860ee981a9e",
"type": "main",
"index": 0
}
]
]
},
"82761634-7472-4ce1-806a-2b80aca985e3": {
"main": [
[
{
"node": "9adbfd4c-bbb9-4c92-bf07-a3b50a92aa02",
"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 - その他, 人工知能
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
音楽生成ワーキングフロー
Suno API、Flux、Runway、Creatomate を使って AI ソングとミュージックビデオを生成
Set
Wait
Merge
+
Set
Wait
Merge
60 ノードJoseph
その他
AIで動くTelegramアシスタント完全入門ガイド（PDF、Brave検索、Google スイート）
Gemini、RAG PDF検索、Google Suiteを用いて多機能Telegramボットを構築
Set
Code
Wait
+
Set
Code
Wait
79 ノードIssam AGGOUR
人工知能
AI自動化とアイグゥチューブの構築と販売
販売AIの自動化とエージェントの構築
If
Set
Wait
+
If
Set
Wait
74 ノードNICKYAI
その他
⚡AI驱动のYouTube播放列表と视频摘要与分析v2
AI YouTube播放列表与视频分析チャットボット
If
Set
Code
+
If
Set
Code
72 ノードdmr
その他
MCPサーバーを基にしたAI Telegramタスクマネージャー
MCP サーバーの統合を备えた AI Telegram タスク マネージメント システム
Set
Switch
Telegram
+
Set
Switch
Telegram
21 ノードFrancis Njenga
その他
AIエージェントを使用してTelegramの食事栄養をGoogle Sheetsに記録
AIエージェントを使用してTelegramから食事の栄養情報をGoogle Sheetsに記録
If
Set
Code
+
If
Set
Code
23 ノードPollupAI
その他
ワークフロー情報
難易度
上級
ノード数22
カテゴリー2
ノードタイプ12
作成者
NanaB
@nanabrownAutomation Developer with 3+ years building innovative solutions on UiPath, Power Platform, and n8n, also with a strong background in both API development and integration. Now specializing in the creation and implementation of AI Agents.
外部リンク
n8n.ioで表示 →
このワークフローを共有