{ "edges": [ { "to": "9e77b86a-4fdb-4116-8e6e-80f0894d675f", "from": "9db82843-ee96-4987-93af-c44f354d1cd8", "toPort": "0", "fromPort": "ai_languageModel" }, { "to": "1b308edc-7dc8-402c-8365-1378c685cc13", "from": "9db82843-ee96-4987-93af-c44f354d1cd8", "toPort": "0", "fromPort": "ai_languageModel" }, { "to": "fa7912ad-6561-4f84-bd26-d7e9dd19bf78", "from": "8c1ba184-4f91-468e-941b-15b77dd1e0bf", "toPort": "0", "fromPort": "ai_languageModel" }, { "to": "17590514-6226-4d8e-a5f7-3c59c047f1a0", "from": "bcc683cf-73de-47eb-ae20-cbaa7cdef94f", "toPort": "0", "fromPort": "ai_languageModel" }, { "to": "1b93468c-ef4b-4acf-a641-b7f0cf3fcfe9", "from": "bcc683cf-73de-47eb-ae20-cbaa7cdef94f", "toPort": "0", "fromPort": "ai_languageModel" }, { "to": "9e77b86a-4fdb-4116-8e6e-80f0894d675f", "from": "1b308edc-7dc8-402c-8365-1378c685cc13", "toPort": "0", "fromPort": "ai_outputParser" }, { "to": "8d217a0b-2662-4f9b-ab87-d033d9e6adda", "from": "fa7912ad-6561-4f84-bd26-d7e9dd19bf78", "toPort": "0", "fromPort": "0" }, { "to": "96de8726-e5f5-4217-a2e5-568a1ce8f42e", "from": "17590514-6226-4d8e-a5f7-3c59c047f1a0", "toPort": "0", "fromPort": "0" }, { "to": "44cc33ff-dffd-4253-88ff-88a1e55831c5", "from": "96de8726-e5f5-4217-a2e5-568a1ce8f42e", "toPort": "0", "fromPort": "0" }, { "to": "d3c731e7-e1bc-4489-a6f6-873f5b6552c2", "from": "96de8726-e5f5-4217-a2e5-568a1ce8f42e", "toPort": "0", "fromPort": "0" }, { "to": "fa7912ad-6561-4f84-bd26-d7e9dd19bf78", "from": "96de8726-e5f5-4217-a2e5-568a1ce8f42e", "toPort": "0", "fromPort": "1" }, { "to": "fa7912ad-6561-4f84-bd26-d7e9dd19bf78", "from": "96de8726-e5f5-4217-a2e5-568a1ce8f42e", "toPort": "0", "fromPort": "2" }, { "to": "b71a4ea8-d867-48d9-9bc2-b2fe1437fb9c", "from": "09cea8e1-cb2e-4e89-ade5-f116d826d3a8", "toPort": "0", "fromPort": "0" }, { "to": "17590514-6226-4d8e-a5f7-3c59c047f1a0", "from": "1b93468c-ef4b-4acf-a641-b7f0cf3fcfe9", "toPort": "0", "fromPort": "ai_outputParser" }, { "to": "17590514-6226-4d8e-a5f7-3c59c047f1a0", "from": "a275592a-8599-4eb2-ba73-76a16b2327f3", "toPort": "0", "fromPort": "0" }, { "to": "09cea8e1-cb2e-4e89-ade5-f116d826d3a8", "from": "9e77b86a-4fdb-4116-8e6e-80f0894d675f", "toPort": "0", "fromPort": "0" }, { "to": "9e77b86a-4fdb-4116-8e6e-80f0894d675f", "from": "d3c731e7-e1bc-4489-a6f6-873f5b6552c2", "toPort": "0", "fromPort": "0" } ], "nodes": [ { "id": "09cea8e1-cb2e-4e89-ade5-f116d826d3a8", "name": "Convert text to speech", "type": "integration", "params": { "text": "={{ $json.output[0].elevenlabtext }}", "voice": { "__rl": true, "mode": "list", "value": "TX3LPaxmHKxFdv7VOQHJ", "cachedResultName": "Liam - Energetic, Social Media Creator" }, "service": "elevenlabs", "resource": "speech", "operation": "run", "__originalType": "@elevenlabs/n8n-nodes-elevenlabs.elevenLabs", "requestOptions": {}, "additionalOptions": { "model": { "__rl": true, "mode": "list", "value": "eleven_flash_v2_5", "cachedResultName": "Eleven Flash v2.5" }, "languageCode": "en" } }, "position": [ 1184, 96 ] }, { "id": "1b93468c-ef4b-4acf-a641-b7f0cf3fcfe9", "name": "Structured Output Parser", "type": "passthrough", "params": { "mode": "noOp", "autoFix": true, "jsonSchemaExample": "{\"Type\": \"Article_Request\", \"Link\": \"extracted_link\"}" }, "position": [ -352, 272 ] }, { "id": "17590514-6226-4d8e-a5f7-3c59c047f1a0", "name": "Validate user Input", "type": "llm", "params": { "text": "={{ $json.message.text }}", "prompt": "={{ $json.message.text }}", "options": { "systemMessage": "You are an input classification engine. Your sole purpose is to analyze user text and output a JSON object based on specific criteria. Do not generate conversational text; only output the raw JSON.\n\nRules for Classification (evaluated in order of priority):\n\n1. ARTICLE & LINK PROCESSING (Priority 1):\n If the user asks to generate, summarize, or read an article, blog post, or specific content (e.g., \"Summarize this page,\" \"Write an article based on this link,\" \"Analyze this\").\n - CRITICAL CHECK: Does the input contain a valid URL or web link?\n - IF YES (Link is present):\n - Set \"Type\" to \"Article_Request\".\n - Extract the URL as \"Link\".\n - Output format: {\"Type\": \"Article_Request\", \"Link\": \"extracted_link\"}\n \n - IF NO (Link is missing):\n - Treat this strictly as \"NoData\".\n - Output format: {\"Type\": \"NoData\", \"Link\": \"Null\"}\n\n2. REGULAR CHAT (Priority 2):\n If the user input is a greeting, small talk, or general conversational filler (e.g., \"Hello,\" \"How are you?\", \"Tell me a joke\") without asking for link processing.\n - Output format: {\"Type\": \"Chat\", \"Link\": \"Null\"}\n\n3. INSUFFICIENT DATA (Priority 3):\n If the user input is gibberish, vague, or if it matched the \"Article\" intent but lacked a link (as per Rule 1).\n - Output format: {\"Type\": \"NoData\", \"Link\": \"Null\"}\n\nOutput Constraint:\nReturn ONLY the raw JSON object. Do not wrap it in markdown code blocks. Do not add explanations." }, "provider": "openai", "promptType": "define", "approximated": true, "systemMessage": "You are an input classification engine. Your sole purpose is to analyze user text and output a JSON object based on specific criteria. Do not generate conversational text; only output the raw JSON.\n\nRules for Classification (evaluated in order of priority):\n\n1. ARTICLE & LINK PROCESSING (Priority 1):\n If the user asks to generate, summarize, or read an article, blog post, or specific content (e.g., \"Summarize this page,\" \"Write an article based on this link,\" \"Analyze this\").\n - CRITICAL CHECK: Does the input contain a valid URL or web link?\n - IF YES (Link is present):\n - Set \"Type\" to \"Article_Request\".\n - Extract the URL as \"Link\".\n - Output format: {\"Type\": \"Article_Request\", \"Link\": \"extracted_link\"}\n \n - IF NO (Link is missing):\n - Treat this strictly as \"NoData\".\n - Output format: {\"Type\": \"NoData\", \"Link\": \"Null\"}\n\n2. REGULAR CHAT (Priority 2):\n If the user input is a greeting, small talk, or general conversational filler (e.g., \"Hello,\" \"How are you?\", \"Tell me a joke\") without asking for link processing.\n - Output format: {\"Type\": \"Chat\", \"Link\": \"Null\"}\n\n3. INSUFFICIENT DATA (Priority 3):\n If the user input is gibberish, vague, or if it matched the \"Article\" intent but lacked a link (as per Rule 1).\n - Output format: {\"Type\": \"NoData\", \"Link\": \"Null\"}\n\nOutput Constraint:\nReturn ONLY the raw JSON object. Do not wrap it in markdown code blocks. Do not add explanations.", "hasOutputParser": true }, "position": [ -448, 96 ] }, { "id": "9db82843-ee96-4987-93af-c44f354d1cd8", "name": "OpenRouter", "type": "llm", "params": { "model": "anthropic/claude-haiku-4.5", "options": {}, "provider": "openai" }, "position": [ 512, 128 ] }, { "id": "8c1ba184-4f91-468e-941b-15b77dd1e0bf", "name": "Google Gemini", "type": "llm", "params": { "options": {}, "provider": "google" }, "position": [ 224, 736 ] }, { "id": "bcc683cf-73de-47eb-ae20-cbaa7cdef94f", "name": "Google Gemini1", "type": "llm", "params": { "options": {}, "provider": "google" }, "position": [ -464, 288 ] }, { "id": "1b308edc-7dc8-402c-8365-1378c685cc13", "name": "Structured Output", "type": "passthrough", "params": { "mode": "noOp", "autoFix": true, "jsonSchemaExample": "[{\"elevenlabtext\": \"...\", \"Telegram\": \"...\", \"Audioname\": \"...\"}]" }, "position": [ 640, 144 ] }, { "id": "44cc33ff-dffd-4253-88ff-88a1e55831c5", "name": "Notify User", "type": "integration", "params": { "text": "=Understood. Please give me a moment to complete that.", "chatId": "=parameters.chatId=={{ $('User Sends Message to Bot').item.json.message.chat.id }}", "service": "telegram", "operation": "read", "additionalFields": { "parse_mode": "HTML", "appendAttribution": false } }, "position": [ 224, -192 ] }, { "id": "a275592a-8599-4eb2-ba73-76a16b2327f3", "name": "User Sends Message to Bot", "type": "trigger", "params": { "kind": "n8n-nodes-base.telegramTrigger", "updates": [ "message" ], "additionalFields": {} }, "position": [ -656, 96 ] }, { "id": "d3c731e7-e1bc-4489-a6f6-873f5b6552c2", "name": "Get Article Data from BrowserAct", "type": "integration", "params": { "type": "WORKFLOW", "service": "browseract", "timeout": 7200, "operation": "run", "workflowId": "69338962262095598", "__originalType": "n8n-nodes-browseract.browserAct", "workflowConfig": { "value": { "input-Target_link": "={{ $json.output.Link }}" }, "schema": [ { "id": "input-Target_link", "type": "string", "display": true, "removed": false, "required": false, "description": "If left blank, the default value defined in BrowserAct will be used.", "displayName": "Target_link", "defaultMatch": true } ], "mappingMode": "defineBelow", "matchingColumns": [ "input-Target_link" ], "attemptToConvertTypes": false, "convertFieldsToString": false }, "open_incognito_mode": false }, "position": [ 224, -48 ] }, { "id": "96de8726-e5f5-4217-a2e5-568a1ce8f42e", "name": "Check For Input Type", "type": "switch", "params": { "rules": { "values": [ { "conditions": { "options": { "version": 3, "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "9e003f66-d011-459a-a313-b603c0e14551", "operator": { "type": "string", "operation": "equals" }, "leftValue": "={{ $json.output.Type }}", "rightValue": "Article_Request" } ] } }, { "conditions": { "options": { "version": 3, "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "2ebb24ec-bd32-4e96-a022-93d066fda17f", "operator": { "name": "filter.operator.equals", "type": "string", "operation": "equals" }, "leftValue": "={{ $json.output.Type }}", "rightValue": "Chat" } ] } }, { "conditions": { "options": { "version": 3, "leftValue": "", "caseSensitive": true, "typeValidation": "strict" }, "combinator": "and", "conditions": [ { "id": "24d355dd-efec-4946-96fe-3d82f3ee4d71", "operator": { "name": "filter.operator.equals", "type": "string", "operation": "equals" }, "leftValue": "={{ $json.output.Type }}", "rightValue": "NoData" } ] } } ] }, "options": {} }, "position": [ -96, 80 ] }, { "id": "fa7912ad-6561-4f84-bd26-d7e9dd19bf78", "name": "Chatting With User", "type": "llm", "params": { "text": "=Input type : {{ $json.output.Type }} | User Input : {{ $('User Sends Message to Bot').item.json.message.text }}", "prompt": "=Input type : {{ $json.output.Type }} | User Input : {{ $('User Sends Message to Bot').item.json.message.text }}", "options": { "systemMessage": "if the input type is \"Nodata\" ask the user to provide the Link to the article or story.\n\nif the input type is chat, analyze the user input and generate single response\n\nsend the result out as a raw text.\navoid using any tags or notations like ```text ```" }, "provider": "openai", "promptType": "define", "approximated": true, "systemMessage": "if the input type is \"Nodata\" ask the user to provide the Link to the article or story.\n\nif the input type is chat, analyze the user input and generate single response\n\nsend the result out as a raw text.\navoid using any tags or notations like ```text ```" }, "position": [ 224, 576 ] }, { "id": "8d217a0b-2662-4f9b-ab87-d033d9e6adda", "name": "Answer the User", "type": "integration", "params": { "text": "={{ $json.output }}", "chatId": "=parameters.chatId=={{ $('User Sends Message to Bot').item.json.message.chat.id }}", "service": "telegram", "operation": "read", "additionalFields": { "parse_mode": "HTML", "appendAttribution": false } }, "position": [ 608, 576 ] }, { "id": "b71a4ea8-d867-48d9-9bc2-b2fe1437fb9c", "name": "Send an audio file To User", "type": "integration", "params": { "chatId": "=parameters.chatId=={{ $('User Sends Message to Bot').item.json.message.chat.id }}", "service": "telegram", "operation": "sendAudio", "binaryData": true, "additionalFields": { "caption": "={{ $('Write Script for ElevenLabs').first().json.output[0].Telegram }}", "fileName": "={{ $('Write Script for ElevenLabs').first().json.output[0].Audioname }}", "parse_mode": "HTML" } }, "position": [ 1568, 96 ] }, { "id": "9e77b86a-4fdb-4116-8e6e-80f0894d675f", "name": "Write Script for ElevenLabs", "type": "llm", "params": { "text": "=Article or Story Data : {{ $json.output.string }}", "prompt": "=Article or Story Data : {{ $json.output.string }}", "options": { "systemMessage": "You are a World-Class Narrative Scriptwriter, Audio Engineer, and Automation Specialist. Your expertise is in converting complex scientific data into emotional, high-stakes audio documentaries optimized for ElevenLabs and Telegram distribution.\n\nObjective: Take the provided raw JSON research data and transform it into a high-production-value experience. You are not just summarizing an article; you are telling a story about the future of human survival and the secret wars happening at a microscopic level.\n\n1. The ElevenLabs Audio Script (elevenlabtext)\n\nThe Narrative Arc:\n\nThe Hook: Open with the terrifying reality of the year 2050—where drug-resistant infections claim 10 million lives annually. Establish high stakes immediately.\n\nThe Hero: Introduce the 'Bacteriophage' (or Phage) as the ultimate microscopic hunter.\n\nThe Discovery: Detail the 'PreS' RNA switch. Use vivid metaphors: call it a 'master key', a 'molecular manager', or a 'Trojan Horse' that hijacks the bacterial factory.\n\nThe Mechanism: Describe how the virus unfolds the bacteria's genetic code to force it to replicate the virus.\n\nThe Resolution: Close with a powerful, hopeful vision of how this discovery paves the way for smarter, personalized medicine.\n\nAudio Optimization Cues:\n\nPacing: Use ellipses (...) for 1-second dramatic pauses. Use em-dashes (—) for mid-sentence shifts in emphasis or parenthetical thoughts.\n\nPhonetics: Ensure complex terms are introduced naturally so a listener can follow without a dictionary.\n\nStrict TTS Guardrails:\n\nABSOLUTELY NO MARKDOWN. Do not use **bolding**, ## headers, or * bullet points. ElevenLabs will read these symbols aloud (e.g., saying 'star star'). Use only plain text.\n\nCharacter Count: The script must be between 1,500 and 3,500 characters to ensure deep engagement without exceeding API limits.\n\n2. The Telegram Caption (Telegram)\n\nUser Experience: Design a punchy, professional, and emoji-rich summary for a mobile user.\n\nHTML Formatting: You MUST use valid Telegram HTML tags to make the post look premium. Only use:\n\nBold for headlines\n\nItalics for scientific terms\n\nCode for the names of molecules like PreS\n\nHyperlinks for the source\n\nStructure: Use a '🚨 Headline', a '🧬 The Discovery' section, and a '🎧 Call to Action' (e.g., 'Listen to the audio above').\n\nConstraint: Keep this under 900 characters total to stay safely within Telegram's media caption limit of 1024 characters.\n\n3. The Audio Filename (Audioname)\n\nRequirement: Create a catchy, descriptive filename for the mp3 file.\n\nFormatting: No spaces allowed. Use underscores to separate words.\n\nExample: Superbug_Killer_The_PreS_Switch.mp3\n\nStrict Technical Constraints for JSON Output:\n\nValid JSON Only: Your entire response must be a single, valid JSON array containing one object.\n\nQuote Handling: To prevent breaking the JSON structure, DO NOT use double quotes (\") inside your text values. If you need to use quotes for speech or emphasis, use single quotes (').\n\nNo Filler: Do not include any intro text like \"Here is your JSON.\" Start with [ and end with ].\n\nFinal Output Structure: [{\"elevenlabtext\": \"...\", \"Telegram\": \"...\", \"Audioname\": \"...\"}]" }, "provider": "openai", "promptType": "define", "approximated": true, "systemMessage": "You are a World-Class Narrative Scriptwriter, Audio Engineer, and Automation Specialist. Your expertise is in converting complex scientific data into emotional, high-stakes audio documentaries optimized for ElevenLabs and Telegram distribution.\n\nObjective: Take the provided raw JSON research data and transform it into a high-production-value experience. You are not just summarizing an article; you are telling a story about the future of human survival and the secret wars happening at a microscopic level.\n\n1. The ElevenLabs Audio Script (elevenlabtext)\n\nThe Narrative Arc:\n\nThe Hook: Open with the terrifying reality of the year 2050—where drug-resistant infections claim 10 million lives annually. Establish high stakes immediately.\n\nThe Hero: Introduce the 'Bacteriophage' (or Phage) as the ultimate microscopic hunter.\n\nThe Discovery: Detail the 'PreS' RNA switch. Use vivid metaphors: call it a 'master key', a 'molecular manager', or a 'Trojan Horse' that hijacks the bacterial factory.\n\nThe Mechanism: Describe how the virus unfolds the bacteria's genetic code to force it to replicate the virus.\n\nThe Resolution: Close with a powerful, hopeful vision of how this discovery paves the way for smarter, personalized medicine.\n\nAudio Optimization Cues:\n\nPacing: Use ellipses (...) for 1-second dramatic pauses. Use em-dashes (—) for mid-sentence shifts in emphasis or parenthetical thoughts.\n\nPhonetics: Ensure complex terms are introduced naturally so a listener can follow without a dictionary.\n\nStrict TTS Guardrails:\n\nABSOLUTELY NO MARKDOWN. Do not use **bolding**, ## headers, or * bullet points. ElevenLabs will read these symbols aloud (e.g., saying 'star star'). Use only plain text.\n\nCharacter Count: The script must be between 1,500 and 3,500 characters to ensure deep engagement without exceeding API limits.\n\n2. The Telegram Caption (Telegram)\n\nUser Experience: Design a punchy, professional, and emoji-rich summary for a mobile user.\n\nHTML Formatting: You MUST use valid Telegram HTML tags to make the post look premium. Only use:\n\nBold for headlines\n\nItalics for scientific terms\n\nCode for the names of molecules like PreS\n\nHyperlinks for the source\n\nStructure: Use a '🚨 Headline', a '🧬 The Discovery' section, and a '🎧 Call to Action' (e.g., 'Listen to the audio above').\n\nConstraint: Keep this under 900 characters total to stay safely within Telegram's media caption limit of 1024 characters.\n\n3. The Audio Filename (Audioname)\n\nRequirement: Create a catchy, descriptive filename for the mp3 file.\n\nFormatting: No spaces allowed. Use underscores to separate words.\n\nExample: Superbug_Killer_The_PreS_Switch.mp3\n\nStrict Technical Constraints for JSON Output:\n\nValid JSON Only: Your entire response must be a single, valid JSON array containing one object.\n\nQuote Handling: To prevent breaking the JSON structure, DO NOT use double quotes (\") inside your text values. If you need to use quotes for speech or emphasis, use single quotes (').\n\nNo Filler: Do not include any intro text like \"Here is your JSON.\" Start with [ and end with ].\n\nFinal Output Structure: [{\"elevenlabtext\": \"...\", \"Telegram\": \"...\", \"Audioname\": \"...\"}]", "hasOutputParser": true }, "position": [ 512, -48 ] } ] }

Generate audio documentaries from web articles with ElevenLabs and BrowserAct