<!\\[CDATA\\[([^\\]]+)\\]\\]><\\/title>/) || item.match(/<title>([^<]+)<\\/title>/);\n if (titleMatch) {\n const itemTitle = titleMatch[1].toLowerCase();\n const words = searchTerms.split(' ').filter(w => w.length > 2);\n let score = 0;\n for (let word of words) { if (itemTitle.includes(word)) score++; }\n if (score > bestScore) { bestScore = score; bestMatch = item; episodeTitle = titleMatch[1]; }\n }\n}\nif (bestMatch) {\n const urlMatch = bestMatch.match(/<enclosure[^>]*url=\"([^\"]+)\"[^>]*type=\"audio/);\n if (urlMatch) { mp3Url = urlMatch[1]; }\n}\nif (!mp3Url) {\n const firstMatch = xmlString.match(/<enclosure[^>]*url=\"([^\"]+)\"[^>]*type=\"audio/);\n mp3Url = firstMatch ? firstMatch[1] : '';\n episodeTitle = 'First episode (no match found)';\n}\nreturn { json: { mp3Url: mp3Url, episodeTitle: episodeTitle, matchScore: bestScore, success: mp3Url ? true : false } };" }, "position": [ -2112, 320 ] }, { "id": "934ac04a-c9da-4ddc-ba9a-882f7b8bef0a", "name": "Fetch RSS Feed", "type": "http", "params": { "url": "={{ $json.rssUrl }}", "options": {} }, "position": [ -2304, 320 ] }, { "id": "c2d71e27-2920-4399-afae-fe22db9eb32f", "name": "Build HTML Email", "type": "code", "params": { "jsCode": "const items = $input.all();\n\nfunction mdToHtml(md) {\n const lines = md.split('\\n');\n const out = [];\n let inList = false;\n for (const line of lines) {\n if (line.startsWith('- ')) {\n if (!inList) { out.push('<ul style=\"margin:8px 0;padding-left:20px\">'); inList = true; }\n const content = line.slice(2)\n .replace(/\\*\\*(.+?)\\*\\*/g, '<strong>$1</strong>')\n .replace(/\\[([^\\]]+)\\]\$([^)]+)\$/g, '<a href=\"$2\" style=\"color:#0066cc\">$1</a>');\n out.push(`<li style=\"margin:4px 0\">${content}</li>`);\n } else {\n if (inList) { out.push('</ul>'); inList = false; }\n if (line.trim() === '') continue;\n const content = line\n .replace(/\\*\\*(.+?)\\*\\*/g, '<strong>$1</strong>')\n .replace(/\\[([^\\]]+)\\]\$([^)]+)\$/g, '<a href=\"$2\" style=\"color:#0066cc\">$1</a>');\n out.push(`<p style=\"margin:6px 0\">${content}</p>`);\n }\n }\n if (inList) out.push('</ul>');\n return out.join('\\n');\n}\n\nlet html = `<!DOCTYPE html>\n<html><head><meta charset=\"UTF-8\"></head>\n<body style=\"font-family:Arial,sans-serif;line-height:1.6;color:#333;max-width:700px;margin:0 auto;padding:20px\">\n<h2 style=\"color:#1a1a1a;border-bottom:2px solid #eee;padding-bottom:10px\">Podcast Summaries</h2>\n<p style=\"color:#888;font-size:13px\">Generated: ${new Date().toLocaleDateString()} · ${items.length} episode(s)</p>\n`;\n\nfor (let i = 0; i < items.length; i++) {\n const item = items[i].json;\n let summary = '';\n if (item.message && item.message.content) { summary = item.message.content; }\n else if (item.choices && item.choices[0] && item.choices[0].message) { summary = item.choices[0].message.content; }\n else if (item.content) { summary = item.content; }\n else { summary = 'Error: Could not extract summary'; }\n html += `<div style=\"border:1px solid #e0e0e0;border-radius:8px;padding:20px;margin:20px 0\">${mdToHtml(summary)}</div>`;\n}\n\nhtml += `</body></html>`;\n\nreturn { json: { html: html } };" }, "position": [ -1360, 320 ] }, { "id": "9fb5b75e-c772-4c14-87dc-2dcd9dd8c0de", "name": "Generate Episode Summary", "type": "llm", "params": { "modelId": { "__rl": true, "mode": "list", "value": "gpt-5-mini", "cachedResultName": "GPT-5-MINI" }, "options": {}, "messages": { "values": [ { "content": "=You are a precise podcast summarization engine. Your only job is to fill in the template below using only information found in the transcript. Do not invent, infer, or embellish.\n\n# Formatting Rules (follow strictly):\n- Output ONLY the filled template. No preamble, no closing remarks, no explanations.\n- Keep every pair of ** exactly as shown. Do not remove or alter any markdown characters.\n- Do not change headings, labels, or section order. Do not add new sections.\n- The **Title** and **URL** lines are pre-filled — copy them exactly as given, do not modify them.\n- Bullets are plain text only: no bold, no nesting, no sub-points. Each must start with \"- \" with no blank lines between them.\n- Do not wrap the URL in any format other than: [URL](URL)\n- If the transcript lacks enough information for a section, write [Insufficient data].\n- The output must be valid Markdown.\n\n# Word Limits:\n- Main Topic: 1 sentence, max 20 words\n- Key Points: 3–5 bullets, each 15–25 words\n- Useful Info: 2–3 sentences, 40–60 words\n- Bottom Line: 1 sentence, max 20 words\n\n# Template:\n**Title**: {{ $('Find Episode MP3 URL').item.json.episodeTitle }}\n**URL**: [{{ $('Parse Input URLs').item.json.apple_url }}]({{ $('Parse Input URLs').item.json.apple_url }})\n**Main Topic:** \n**Key Points:**\n- \n**Useful Info:** \n**Bottom Line:** \n\n# Transcript:\n{{ $json.text }}" } ] }, "provider": "openai", "simplify": false }, "position": [ -1696, 320 ] }, { "id": "db9a8c06-c22f-4ddc-9d13-f3fa37c985d6", "name": "Send Summary Email", "type": "integration", "params": { "sendTo": "user@example.com", "message": "={{ $json.html }}", "options": { "appendAttribution": false }, "service": "gmail", "subject": "Podcast Summary", "operation": "read" }, "position": [ -1168, 320 ] } ] }</article></body></html>

{ "edges": [ { "to": "376cb838-552b-4e44-bfd4-99b4285f6918", "from": "934ac04a-c9da-4ddc-ba9a-882f7b8bef0a", "toPort": "0", "fromPort": "0" }, { "to": "db9a8c06-c22f-4ddc-9d13-f3fa37c985d6", "from": "c2d71e27-2920-4399-afae-fe22db9eb32f", "toPort": "0", "fromPort": "0" }, { "to": "cd36a002-1b68-49b6-a2f9-f56c0bbf675d", "from": "b263c0b1-797d-4197-b6aa-4b967d9af20e", "toPort": "0", "fromPort": "0" }, { "to": "b263c0b1-797d-4197-b6aa-4b967d9af20e", "from": "fca607d8-ec3a-434f-96ea-b9f3151f9ba9", "toPort": "0", "fromPort": "0" }, { "to": "934ac04a-c9da-4ddc-ba9a-882f7b8bef0a", "from": "970c1467-b612-4460-8c86-d3643f617bcf", "toPort": "0", "fromPort": "0" }, { "to": "elevenlabs-stt", "from": "376cb838-552b-4e44-bfd4-99b4285f6918", "toPort": "0", "fromPort": "0" }, { "to": "c2d71e27-2920-4399-afae-fe22db9eb32f", "from": "9fb5b75e-c772-4c14-87dc-2dcd9dd8c0de", "toPort": "0", "fromPort": "0" }, { "to": "970c1467-b612-4460-8c86-d3643f617bcf", "from": "cd36a002-1b68-49b6-a2f9-f56c0bbf675d", "toPort": "0", "fromPort": "0" }, { "to": "9fb5b75e-c772-4c14-87dc-2dcd9dd8c0de", "from": "elevenlabs-stt", "toPort": "0", "fromPort": "0" } ], "nodes": [ { "id": "elevenlabs-stt", "name": "Transcribe Episode with ElevenLabs", "type": "http", "params": { "url": "https://api.elevenlabs.io/v1/speech-to-text", "method": "POST", "options": { "timeout": 600000 }, "sendBody": true, "contentType": "multipart-form-data", "authentication": "genericCredentialType", "bodyParameters": { "parameters": [ { "name": "cloud_storage_url", "value": "={{ $json.mp3Url }}" }, { "name": "model_id", "value": "scribe_v2" } ] }, "genericAuthType": "httpHeaderAuth" }, "position": [ -1872, 320 ] }, { "id": "fca607d8-ec3a-434f-96ea-b9f3151f9ba9", "name": "Submit Podcast URLs", "type": "trigger", "params": { "kind": "n8n-nodes-base.formTrigger", "options": { "appendAttribution": false }, "formTitle": "Apple Podcasts Summarizer", "formFields": { "values": [ { "fieldType": "textarea", "fieldLabel": "Input Apple Podcast links", "placeholder": "https://podcasts.apple.com/mn/podcast/seasonality-isnt-a-problem-its-a-profit-opportunity-ep-952/id1254720112?i=1+1234567890", "requiredField": true } ] } }, "position": [ -3104, 320 ] }, { "id": "b263c0b1-797d-4197-b6aa-4b967d9af20e", "name": "Parse Input URLs", "type": "code", "params": { "jsCode": "const input = $input.first().json['Input Apple Podcast links'];\nconst urls = input.split('\\n').filter(u => u.trim() !== '');\nreturn urls.map(url => {\n const match = url.match(/id(\\d+)/);\n return {\n json: {\n apple_url: url.trim(),\n podcastId: match ? match[1] : ''\n }\n };\n});" }, "position": [ -2880, 320 ] }, { "id": "cd36a002-1b68-49b6-a2f9-f56c0bbf675d", "name": "Look Up RSS Feed via iTunes", "type": "http", "params": { "url": "=https://itunes.apple.com/lookup?id={{ $json.podcastId }}&entity=podcast", "options": {} }, "position": [ -2688, 320 ] }, { "id": "970c1467-b612-4460-8c86-d3643f617bcf", "name": "Extract RSS Feed URL", "type": "code", "params": { "mode": "runOnceForEachItem", "jsCode": "const rawData = $input.item.json;\nlet data;\nif (rawData.data !== undefined) {\n data = typeof rawData.data === 'string' ? JSON.parse(rawData.data) : rawData.data;\n} else {\n data = rawData;\n}\nlet rssUrl = '';\nif (data.results && data.results.length > 0) { rssUrl = data.results[0].feedUrl; }\nreturn { json: { rssUrl: rssUrl } };" }, "position": [ -2496, 320 ] }, { "id": "376cb838-552b-4e44-bfd4-99b4285f6918", "name": "Find Episode MP3 URL", "type": "code", "params": { "mode": "runOnceForEachItem", "jsCode": "const xml = $input.item.json.data;\nconst xmlString = typeof xml === 'string' ? xml : JSON.stringify(xml);\nconst originalUrl = $('Parse Input URLs').item.json.apple_url;\nconst urlParts = originalUrl.split('/');\nconst urlSlug = urlParts.find(part => part.includes('-') && part.length > 10) || '';\nconst searchTerms = urlSlug.toLowerCase().replace(/-/g, ' ');\nconst items = xmlString.split('');\nlet mp3Url = '';\nlet episodeTitle = '';\nlet bestMatch = null;\nlet bestScore = 0;\nfor (let item of items) {\n const titleMatch = item.match(/<!\\[CDATA\\[([^\\]]+)\\]\\]><\\/title>/) || item.match(/<title>([^<]+)<\\/title>/);\n if (titleMatch) {\n const itemTitle = titleMatch[1].toLowerCase();\n const words = searchTerms.split(' ').filter(w => w.length > 2);\n let score = 0;\n for (let word of words) { if (itemTitle.includes(word)) score++; }\n if (score > bestScore) { bestScore = score; bestMatch = item; episodeTitle = titleMatch[1]; }\n }\n}\nif (bestMatch) {\n const urlMatch = bestMatch.match(/<enclosure[^>]*url=\"([^\"]+)\"[^>]*type=\"audio/);\n if (urlMatch) { mp3Url = urlMatch[1]; }\n}\nif (!mp3Url) {\n const firstMatch = xmlString.match(/<enclosure[^>]*url=\"([^\"]+)\"[^>]*type=\"audio/);\n mp3Url = firstMatch ? firstMatch[1] : '';\n episodeTitle = 'First episode (no match found)';\n}\nreturn { json: { mp3Url: mp3Url, episodeTitle: episodeTitle, matchScore: bestScore, success: mp3Url ? true : false } };" }, "position": [ -2112, 320 ] }, { "id": "934ac04a-c9da-4ddc-ba9a-882f7b8bef0a", "name": "Fetch RSS Feed", "type": "http", "params": { "url": "={{ $json.rssUrl }}", "options": {} }, "position": [ -2304, 320 ] }, { "id": "c2d71e27-2920-4399-afae-fe22db9eb32f", "name": "Build HTML Email", "type": "code", "params": { "jsCode": "const items = $input.all();\n\nfunction mdToHtml(md) {\n const lines = md.split('\\n');\n const out = [];\n let inList = false;\n for (const line of lines) {\n if (line.startsWith('- ')) {\n if (!inList) { out.push('<ul style=\"margin:8px 0;padding-left:20px\">'); inList = true; }\n const content = line.slice(2)\n .replace(/\\*\\*(.+?)\\*\\*/g, '<strong>$1</strong>')\n .replace(/\\[([^\\]]+)\\]\$([^)]+)\$/g, '<a href=\"$2\" style=\"color:#0066cc\">$1</a>');\n out.push(`<li style=\"margin:4px 0\">${content}</li>`);\n } else {\n if (inList) { out.push('</ul>'); inList = false; }\n if (line.trim() === '') continue;\n const content = line\n .replace(/\\*\\*(.+?)\\*\\*/g, '<strong>$1</strong>')\n .replace(/\\[([^\\]]+)\\]\$([^)]+)\$/g, '<a href=\"$2\" style=\"color:#0066cc\">$1</a>');\n out.push(`<p style=\"margin:6px 0\">${content}</p>`);\n }\n }\n if (inList) out.push('</ul>');\n return out.join('\\n');\n}\n\nlet html = `<!DOCTYPE html>\n<html><head><meta charset=\"UTF-8\"></head>\n<body style=\"font-family:Arial,sans-serif;line-height:1.6;color:#333;max-width:700px;margin:0 auto;padding:20px\">\n<h2 style=\"color:#1a1a1a;border-bottom:2px solid #eee;padding-bottom:10px\">Podcast Summaries</h2>\n<p style=\"color:#888;font-size:13px\">Generated: ${new Date().toLocaleDateString()} · ${items.length} episode(s)</p>\n`;\n\nfor (let i = 0; i < items.length; i++) {\n const item = items[i].json;\n let summary = '';\n if (item.message && item.message.content) { summary = item.message.content; }\n else if (item.choices && item.choices[0] && item.choices[0].message) { summary = item.choices[0].message.content; }\n else if (item.content) { summary = item.content; }\n else { summary = 'Error: Could not extract summary'; }\n html += `<div style=\"border:1px solid #e0e0e0;border-radius:8px;padding:20px;margin:20px 0\">${mdToHtml(summary)}</div>`;\n}\n\nhtml += `</body></html>`;\n\nreturn { json: { html: html } };" }, "position": [ -1360, 320 ] }, { "id": "9fb5b75e-c772-4c14-87dc-2dcd9dd8c0de", "name": "Generate Episode Summary", "type": "llm", "params": { "modelId": { "__rl": true, "mode": "list", "value": "gpt-5-mini", "cachedResultName": "GPT-5-MINI" }, "options": {}, "messages": { "values": [ { "content": "=You are a precise podcast summarization engine. Your only job is to fill in the template below using only information found in the transcript. Do not invent, infer, or embellish.\n\n# Formatting Rules (follow strictly):\n- Output ONLY the filled template. No preamble, no closing remarks, no explanations.\n- Keep every pair of ** exactly as shown. Do not remove or alter any markdown characters.\n- Do not change headings, labels, or section order. Do not add new sections.\n- The **Title** and **URL** lines are pre-filled — copy them exactly as given, do not modify them.\n- Bullets are plain text only: no bold, no nesting, no sub-points. Each must start with \"- \" with no blank lines between them.\n- Do not wrap the URL in any format other than: [URL](URL)\n- If the transcript lacks enough information for a section, write [Insufficient data].\n- The output must be valid Markdown.\n\n# Word Limits:\n- Main Topic: 1 sentence, max 20 words\n- Key Points: 3–5 bullets, each 15–25 words\n- Useful Info: 2–3 sentences, 40–60 words\n- Bottom Line: 1 sentence, max 20 words\n\n# Template:\n**Title**: {{ $('Find Episode MP3 URL').item.json.episodeTitle }}\n**URL**: [{{ $('Parse Input URLs').item.json.apple_url }}]({{ $('Parse Input URLs').item.json.apple_url }})\n**Main Topic:** \n**Key Points:**\n- \n**Useful Info:** \n**Bottom Line:** \n\n# Transcript:\n{{ $json.text }}" } ] }, "provider": "openai", "simplify": false }, "position": [ -1696, 320 ] }, { "id": "db9a8c06-c22f-4ddc-9d13-f3fa37c985d6", "name": "Send Summary Email", "type": "integration", "params": { "sendTo": "user@example.com", "message": "={{ $json.html }}", "options": { "appendAttribution": false }, "service": "gmail", "subject": "Podcast Summary", "operation": "read" }, "position": [ -1168, 320 ] } ] }</article></body></html>

Summarize Apple podcast episodes with ElevenLabs and GPT-5-MINI