This workflow follows the Gmail Trigger → HTTP Request recipe pattern — see all workflows that pair these two integrations.
The workflow JSON
Copy or download the full n8n JSON below. Paste it into a new n8n workflow, add your credentials, activate. Full import guide →
{
"id": "documents-poc",
"name": "Documents - Extraction Assistant",
"active": false,
"settings": {
"executionOrder": "v1"
},
"nodes": [
{
"parameters": {
"pollTimes": {
"item": [
{
"mode": "everyMinute"
}
]
},
"simple": false,
"filters": {
"q": "has:attachment",
"readStatus": "unread"
},
"options": {
"dataPropertyAttachmentsPrefixName": "attachment_",
"downloadAttachments": true
}
},
"id": "10bcbee4-94f3-4485-8bf0-4419e921b6e6",
"name": "Gmail Trigger",
"type": "n8n-nodes-base.gmailTrigger",
"typeVersion": 1.4,
"position": [
240,
304
],
"credentials": {
"gmailOAuth2": {
"name": "<your credential>"
}
}
},
{
"parameters": {
"operation": "binaryToPropery",
"binaryPropertyName": "attachment_0",
"destinationKey": "fileBase64",
"options": {}
},
"id": "511dcca8-19eb-4fbd-8cde-6975e658dc82",
"name": "Extract Attachment",
"type": "n8n-nodes-base.extractFromFile",
"typeVersion": 1.1,
"position": [
464,
304
]
},
{
"parameters": {
"jsCode": "const base64Data = $json.fileBase64;\nconst triggerItem = $('Gmail Trigger').item;\nconst binaryMeta = triggerItem.binary && triggerItem.binary['attachment_0'];\nconst mimeType = (binaryMeta && binaryMeta.mimeType) || 'application/pdf';\nconst fileName = (binaryMeta && binaryMeta.fileName) || 'attachment';\nconst email = triggerItem.json;\n\nconst isPdf = mimeType.includes('pdf');\nconst contentBlock = isPdf\n ? { type: 'document', source: { type: 'base64', media_type: 'application/pdf', data: base64Data } }\n : { type: 'image', source: { type: 'base64', media_type: mimeType, data: base64Data } };\n\nconst systemPrompt = `You extract structured data from business documents (invoices, receipts, forms).\n\nLook at the attached document and return ONLY a JSON object, no markdown, no code fences, in exactly this shape:\n{\n \"document_type\": \"invoice\" | \"receipt\" | \"form\" | \"unknown\",\n \"vendor\": \"<string or null>\",\n \"total_amount\": <number or null>,\n \"document_date\": \"<YYYY-MM-DD or null>\",\n \"fields\": [ { \"name\": \"<field name>\", \"value\": \"<extracted value>\", \"page\": <page number the value was found on, integer, 1-indexed> } ],\n \"confident\": true or false,\n \"reason\": \"<one short sentence, for internal use only>\"\n}\n\nOnly set confident to true if the document is clearly legible and the key fields (vendor, total, date) were actually found. If the document is blank, unreadable, or not actually a business document, set confident to false and explain why in reason.`;\n\nconst requestBody = {\n model: 'claude-sonnet-5',\n max_tokens: 1536,\n system: systemPrompt,\n messages: [\n { role: 'user', content: [ contentBlock, { type: 'text', text: 'Extract the structured data from this document.' } ] }\n ],\n};\n\nlet fromAddress = '';\nif (email.from) {\n if (typeof email.from === 'string') {\n const m = email.from.match(/<([^>]+)>/);\n fromAddress = m ? m[1] : email.from;\n } else if (typeof email.from === 'object') {\n fromAddress = (email.from.value && email.from.value[0] && email.from.value[0].address) || email.from.text || '';\n }\n}\n\nreturn [{\n json: {\n requestBody,\n original: {\n fileName,\n mimeType,\n base64Data,\n emailSubject: email.subject,\n emailFrom: fromAddress,\n messageId: email.messageId || email.id,\n },\n },\n}];"
},
"id": "3439c4dd-f9ef-4745-b61a-2405de7576ef",
"name": "Build Extraction Prompt",
"type": "n8n-nodes-base.code",
"typeVersion": 2,
"position": [
688,
304
]
},
{
"parameters": {
"method": "POST",
"url": "https://api.anthropic.com/v1/messages",
"authentication": "predefinedCredentialType",
"nodeCredentialType": "anthropicApi",
"sendHeaders": true,
"headerParameters": {
"parameters": [
{
"name": "anthropic-version",
"value": "2023-06-01"
}
]
},
"sendBody": true,
"specifyBody": "json",
"jsonBody": "={{ $json.requestBody }}",
"options": {}
},
"id": "604f1a9b-c94b-43c6-8d3a-b5d7ef54669d",
"name": "Call Claude",
"type": "n8n-nodes-base.httpRequest",
"typeVersion": 4.2,
"position": [
912,
304
],
"credentials": {
"anthropicApi": {
"name": "<your credential>"
}
}
},
{
"parameters": {
"jsCode": "const response = $input.first().json;\nconst original = $('Build Extraction Prompt').item.json.original;\n\nlet text = '';\nif (response.content && response.content[0] && response.content[0].text) {\n text = response.content[0].text;\n}\ntext = text.trim();\nif (text.startsWith('```')) {\n text = text.replace(/^```(json)?/i, '').replace(/```$/, '').trim();\n}\n\nlet parsed;\ntry {\n parsed = JSON.parse(text);\n} catch (e) {\n parsed = { document_type: 'unknown', vendor: null, total_amount: null, document_date: null, fields: [], confident: false, reason: 'Failed to parse model response as JSON' };\n}\n\nconst confident = parsed.confident === true;\nconst status = confident ? 'success' : 'needs_review';\n\nconst documentRecord = {\n source: 'gmail-attachment',\n sourceRef: `${original.messageId}:${original.fileName}`,\n fileName: original.fileName,\n fileBase64: original.base64Data,\n documentType: parsed.document_type || 'unknown',\n vendor: parsed.vendor || null,\n totalAmount: parsed.total_amount || null,\n documentDate: parsed.document_date || null,\n extracted: { fields: parsed.fields || [], emailSubject: original.emailSubject, emailFrom: original.emailFrom },\n status,\n reason: parsed.reason || null,\n};\n\nconst audit = {\n source: 'documents',\n actor: original.emailFrom || 'unknown',\n action: 'extract_document',\n details: { fileName: original.fileName, documentType: documentRecord.documentType, status },\n status,\n reason: parsed.reason || null,\n};\n\nreturn [{ json: { documentRecord, audit } }];"
},
"id": "62c44cfa-daaf-4a8a-807b-dae7627a6d74",
"name": "Parse Extraction",
"type": "n8n-nodes-base.code",
"typeVersion": 2,
"position": [
1120,
304
]
},
{
"parameters": {
"jsCode": "return [{ json: $input.first().json.audit }];"
},
"id": "5ea54a94-d8b5-49c3-b2f9-25e066d99e82",
"name": "Build Audit Record",
"type": "n8n-nodes-base.code",
"typeVersion": 2,
"position": [
1344,
208
]
},
{
"parameters": {
"operation": "executeQuery",
"query": "INSERT INTO audit_log (source, actor, action, details, status, reason) VALUES ($1, $2, $3, $4, $5, $6)",
"options": {
"queryReplacement": "={{ [$json.source, $json.actor, $json.action, $json.details ? JSON.stringify($json.details) : null, $json.status, $json.reason] }}"
}
},
"id": "a4ba1d86-2660-4723-a068-24e7dc89facf",
"name": "Log Audit",
"type": "n8n-nodes-base.postgres",
"typeVersion": 2.6,
"position": [
1568,
208
],
"credentials": {
"postgres": {
"name": "<your credential>"
}
}
},
{
"parameters": {
"jsCode": "return [{ json: $input.first().json.documentRecord }];"
},
"id": "8a813375-0f3c-4440-ac82-829db5202252",
"name": "Build Document Record",
"type": "n8n-nodes-base.code",
"typeVersion": 2,
"position": [
1344,
400
]
},
{
"parameters": {
"operation": "executeQuery",
"query": "INSERT INTO documents (source, source_ref, file_name, file_bytes, document_type, vendor, total_amount, document_date, extracted, status, reason) VALUES ($1, $2, $3, decode($4, 'base64'), $5, $6, $7, $8, $9, $10, $11)",
"options": {
"queryReplacement": "={{ [$json.source, $json.sourceRef, $json.fileName, $json.fileBase64, $json.documentType, $json.vendor, $json.totalAmount, $json.documentDate, JSON.stringify($json.extracted), $json.status, $json.reason] }}"
}
},
"id": "d8d5312c-964f-46c4-82e6-fe1222da7610",
"name": "Insert Document",
"type": "n8n-nodes-base.postgres",
"typeVersion": 2.6,
"position": [
1568,
400
],
"credentials": {
"postgres": {
"name": "<your credential>"
}
}
}
],
"connections": {
"Gmail Trigger": {
"main": [
[
{
"node": "Extract Attachment",
"type": "main",
"index": 0
}
]
]
},
"Extract Attachment": {
"main": [
[
{
"node": "Build Extraction Prompt",
"type": "main",
"index": 0
}
]
]
},
"Build Extraction Prompt": {
"main": [
[
{
"node": "Call Claude",
"type": "main",
"index": 0
}
]
]
},
"Call Claude": {
"main": [
[
{
"node": "Parse Extraction",
"type": "main",
"index": 0
}
]
]
},
"Parse Extraction": {
"main": [
[
{
"node": "Build Audit Record",
"type": "main",
"index": 0
},
{
"node": "Build Document Record",
"type": "main",
"index": 0
}
]
]
},
"Build Audit Record": {
"main": [
[
{
"node": "Log Audit",
"type": "main",
"index": 0
}
]
]
},
"Build Document Record": {
"main": [
[
{
"node": "Insert Document",
"type": "main",
"index": 0
}
]
]
}
}
}
Credentials you'll need
Each integration node will prompt for credentials when you import. We strip credential IDs before publishing — you'll add your own.
anthropicApigmailOAuth2postgres
Pro
For the full experience including quality scoring and batch install features for each workflow upgrade to Pro
About this workflow
Documents - Extraction Assistant. Uses gmailTrigger, httpRequest, postgres. Event-driven trigger; 9 nodes.
Source: https://github.com/JamesSoria/n8n-automation-poc/blob/main/n8n-workflows/documents-extraction-assistant.json — original creator credit. Request a take-down →
Related workflows
Workflows that share integrations, category, or trigger type with this one. All free to copy and import.
gbrain - Gmail Receiver (broad + classifier). Uses gmailTrigger, postgres, httpRequest. Event-driven trigger; 5 nodes.
Gmail Trigger, Postgres, HTTP Request
This workflow ingests proof-of-delivery and completion documents from Gmail or a webhook, extracts key fields with an OpenRouter vision model, reconciles them against Google Sheets dispatch data, arch
Gmail Trigger, HTTP Request, Google Sheets +2
Limit. Uses gmailTrigger, httpRequest, limit, respondToWebhook. Event-driven trigger; 40 nodes.
Gmail Trigger, HTTP Request
This workflow is ideal for IT professionals, security analysts, and organizations looking to enhance their email security practices. It is particularly useful for those who need to analyze Gmail email
Gmail Trigger, HTTP Request
Gmailtrigger Workflow. Uses gmailTrigger, httpRequest. Event-driven trigger; 40 nodes.
Gmail Trigger, HTTP Request