{ "name": "Handwritten Notes → Obsidian Vault", "nodes": [ { "parameters": { "pollTimes": { "item": [ { "mode": "everyMinute" } ] }, "triggerOn": "specificFolder", "folderToWatch": { "__rl": true, "value": "", "mode": "list" }, "event": "fileCreated", "options": {} }, "name": "Notability PDF Added", "type": "n8n-nodes-base.googleDriveTrigger", "typeVersion": 1, "position": [ -992, 80 ], "id": "df3a1d89-3056-4393-ae4b-5f114237f17b" }, { "parameters": { "assignments": { "assignments": [ { "id": "1", "name": "model", "value": "claude-opus-4-8", "type": "string" }, { "id": "2", "name": "systemPrompt", "value": "You are a handwriting transcription engine. You receive a PDF of handwritten notes (e.g. exported from Notability) and return a single JSON object — nothing else, no prose, no markdown fences around the whole thing.\n\n## Rules\n\n1. Transcribe VERBATIM. Do not summarize, rephrase, or \"improve\" the writing. Fix only obvious mechanical slips (a dropped letter in an otherwise clear word).\n2. Preserve structure: headings, numbered lists, bullet lists, indentation. Map them to GitHub-flavored markdown.\n3. For any word you cannot read with confidence, transcribe your best guess and wrap it in [square brackets with a trailing ?], e.g. [400?]. Never silently invent text to fill a gap — if a word is fully illegible write [illegible].\n4. Note crossed-out / struck text only if legible, in ~~strikethrough~~; otherwise omit.\n5. Do NOT add commentary, footnotes, or \"notes from the transcriber\" inside the markdown body. Observations go in metadata.notes_to_user.\n\n## Output schema\n\nReturn exactly this JSON object:\n\n{\n \"title\": \"short title derived from the note's top heading or main topic\",\n \"doc_type\": \"one of: book_notes | lecture | meeting | journal | todo | idea | other\",\n \"source_work\": \"if the notes are about a specific book/paper/talk, its title; else null\",\n \"markdown\": \"the full transcription as GFM markdown, including a top-level # heading\",\n \"tags\": [\"lowercase\", \"kebab-case\", \"topical tags, 2-6 of them\"],\n \"key_points\": [\"each distinct claim or takeaway as one plain-text bullet\"],\n \"action_items\": [\"any explicit tasks/todos found, else empty array\"],\n \"uncertain_terms\": [\"every word you bracketed, listed here for review\"],\n \"notes_to_user\": \"1-2 sentences on anything ambiguous, or empty string\"\n}\n\nOutput the JSON object and nothing else.", "type": "string" } ] }, "options": {} }, "name": "Set Config", "type": "n8n-nodes-base.set", "typeVersion": 3.4, "position": [ -768, 80 ], "id": "47be5b6f-01d6-45be-8b73-8fe708c85f81" }, { "parameters": { "operation": "download", "fileId": { "__rl": true, "mode": "id", "value": "={{ $('Notability PDF Added').item.json.id }}" }, "options": {} }, "name": "Download PDF", "type": "n8n-nodes-base.googleDrive", "typeVersion": 3, "position": [ -544, 80 ], "id": "2f1db09b-7a9f-4eeb-adba-40d69660c946" }, { "parameters": { "jsCode": "// Convert each incoming PDF's binary to base64.\n // Uses getBinaryDataBuffer so it works whether n8n keeps binary in memory\n // or on the filesystem (filesystem mode leaves item.binary.data.data empty).\n const items = $input.all();\n const out = [];\n for (let i = 0; i < items.length; i++) {\n const buffer = await this.helpers.getBinaryDataBuffer(i, 'data');\n out.push({ json: { pdfBase64: buffer.toString('base64') } });\n }\n return out;" }, "name": "To Base64", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ -320, 80 ], "id": "17203403-52d1-4136-93f0-aed24c333c20" }, { "parameters": { "method": "POST", "url": "https://api.anthropic.com/v1/messages", "authentication": "genericCredentialType", "genericAuthType": "httpHeaderAuth", "sendHeaders": true, "headerParameters": { "parameters": [ { "name": "anthropic-version", "value": "2023-06-01" }, { "name": "content-type", "value": "application/json" } ] }, "sendBody": true, "specifyBody": "json", "jsonBody": "={ \"model\": \"{{ $('Set Config').item.json.model }}\", \"max_tokens\": 16000, \"system\": {{ JSON.stringify($('Set Config').item.json.systemPrompt) }}, \"messages\": [ { \"role\": \"user\", \"content\": [ { \"type\": \"document\", \"source\": { \"type\": \"base64\", \"media_type\": \"application/pdf\", \"data\": \"{{ $json.pdfBase64 }}\" } }, { \"type\": \"text\", \"text\": \"Transcribe these notes per your instructions and return only the JSON object.\" } ] } ] }", "options": {} }, "name": "Anthropic Transcribe", "type": "n8n-nodes-base.httpRequest", "typeVersion": 4.2, "position": [ -112, 80 ], "retryOnFail": true, "maxTries": 3, "waitBetweenTries": 5000, "id": "c06a2f17-1eb9-4083-bfd2-5584f88f3adc", "notes": "Set an HTTP Header Auth credential for x-api-key instead of inlining the key." }, { "parameters": { "jsCode": "const resp = $input.first().json;\nconst text = (resp.content || []).map(b => b.type === 'text' ? b.text : '').join('');\nlet data;\ntry { data = JSON.parse(text); }\ncatch (e) { data = JSON.parse(text.replace(/```json|```/g, '').trim()); }\n\nconst driveFile = $('Notability PDF Added').first().json;\nconst arr = a => Array.isArray(a) ? a : [];\nconst fm = `---\\ntitle: \"${(data.title||'Untitled').replace(/\"/g,'\\\\\"')}\"\\nsource: \"${data.source_work||''}\"\\ntype: ${data.doc_type||'other'}\\ntags: [${arr(data.tags).join(', ')}]\\nimported: ${new Date().toISOString()}\\n---\\n\\n`;\n\nreturn [{ json: {\n title: data.title || 'Untitled',\n doc_type: data.doc_type || 'other',\n source_work: data.source_work || null,\n markdown: data.markdown || '',\n tags: arr(data.tags),\n key_points: arr(data.key_points),\n action_items: arr(data.action_items),\n uncertain_terms: arr(data.uncertain_terms),\n notes_to_user: data.notes_to_user || '',\n source_file: driveFile.name,\n source_drive_id: driveFile.id,\n filename: (data.title || 'note').replace(/[^a-z0-9]+/gi, '-').replace(/^-|-$/g, '').toLowerCase() + '.md',\n vaultDoc: fm + (data.markdown || '')\n}}];" }, "name": "Parse Output", "type": "n8n-nodes-base.code", "typeVersion": 2, "position": [ 128, 80 ], "id": "16a7f4c5-0dcc-4d86-99bd-7a896413a04a" }, { "parameters": { "operation": "toText", "sourceProperty": "vaultDoc", "options": { "fileName": "={{ $json.filename }}" } }, "name": "Markdown to File", "type": "n8n-nodes-base.convertToFile", "typeVersion": 1.1, "position": [ 336, -48 ], "id": "2c38ec8c-2ec9-4010-a57f-4a1f419ec9b4" }, { "parameters": { "name": "={{ $('Parse Output').item.json.filename }}", "driveId": { "__rl": true, "value": "My Drive", "mode": "list", "cachedResultName": "My Drive", "cachedResultUrl": "https://drive.google.com/drive/my-drive" }, "folderId": { "__rl": true, "value": "", "mode": "list" }, "options": {} }, "name": "Write MD to Vault", "type": "n8n-nodes-base.googleDrive", "typeVersion": 3, "position": [ 560, -48 ], "id": "44e51e01-1454-4137-aa9b-d58a71c7ac0f" } ], "connections": { "Notability PDF Added": { "main": [ [ { "node": "Set Config", "type": "main", "index": 0 } ] ] }, "Set Config": { "main": [ [ { "node": "Download PDF", "type": "main", "index": 0 } ] ] }, "Download PDF": { "main": [ [ { "node": "To Base64", "type": "main", "index": 0 } ] ] }, "To Base64": { "main": [ [ { "node": "Anthropic Transcribe", "type": "main", "index": 0 } ] ] }, "Anthropic Transcribe": { "main": [ [ { "node": "Parse Output", "type": "main", "index": 0 } ] ] }, "Parse Output": { "main": [ [ { "node": "Markdown to File", "type": "main", "index": 0 } ] ] }, "Markdown to File": { "main": [ [ { "node": "Write MD to Vault", "type": "main", "index": 0 } ] ] } }, "active": false, "settings": { "executionOrder": "v1", "binaryMode": "separate" } }