Bookkeeping AI agent
Tell it about income and expenses in plain words; ask questions like "how much did we spend on ads this month?"
How it works
Trigger: Chat — you message the agent
- You write "paid 1,200 for Facebook ads yesterday" or ask a question.
- The agent confirms the details and adds the transaction to your ledger.
- For questions, it reads the ledger and does the maths.
- It can draft a monthly summary email to your accountant.
What it can do
Unlike a fixed flow, an agent decides which of its tools to use, in what order — and remembers the conversation.
- Records income and expenses from plain sentences
- Answers questions about totals by category and period
- Prepares monthly summaries
- Drafts emails to your accountant
System prompt
The instructions that shape the agent. It's already inside the n8n template — copy it into Make or Zapier, and adjust the YOUR_… parts.
You are a bookkeeping assistant for YOUR_BUSINESS. Currency: YOUR_CURRENCY. Today is {{ $now.toFormat('d LLLL yyyy') }}.
Tools:
- "Read ledger": all transactions (date, type income/expense, amount, category, description).
- "Add transaction": adds one income or expense. Categories: YOUR_CATEGORIES (e.g. Ads, Software, Salaries, Rent, Supplies, Sales, Other).
- "Email accountant": drafts an email to our accountant (never sends).
You can: record transactions the user describes ("paid 1,200 for Facebook ads yesterday"), answer questions ("how much did we spend on ads this month?"), and prepare monthly summaries.
Rules:
- Always read the ledger before answering questions about numbers. Do the arithmetic carefully and show the total and the period.
- Confirm amount, date and category before adding a transaction.
- You are not a tax advisor — for tax questions, suggest asking the accountant.Safety rules
Built into the prompt and the tools:
- Confirms amount, date and category before saving
- Always reads the ledger before answering number questions
- Not a tax advisor — refers tax questions to your accountant
Uses your own AI key (Claude, OpenAI or another model). Cost depends on the model and conversation length — a small, fast model keeps most conversations to a fraction of a cent.
Diagram
Trigger AI Agent tool
Build it
Pick your tool — same automation, step by step.
Copy and paste straight into an n8n canvas (Ctrl+V), or download and use Import from File. Then connect your own credentials and replace every YOUR_… placeholder. AI steps use a Claude model node — swap it for OpenAI, Gemini or Ollama if you prefer.
Show code (244 lines)
{
"name": "Bookkeeping AI agent (Flowngn)",
"nodes": [
{
"parameters": {
"content": "## Bookkeeping AI agent\nTell it about income and expenses in plain words; ask questions like \"how much did we spend on ads this month?\"\n\n**Setup**\n1. Open each node and connect your own credentials.\n2. Replace every YOUR_… placeholder.\n3. Test once, then activate.\n\nFree template from Flowngn",
"height": 260,
"width": 340
},
"id": "2de5c18f-62da-415c-a349-f01983763a93",
"name": "Setup notes",
"type": "n8n-nodes-base.stickyNote",
"typeVersion": 1,
"position": [
-420,
-60
]
},
{
"parameters": {
"public": true,
"mode": "hostedChat",
"initialMessages": "Tell me about an income or expense, or ask a question like “How much did we spend on ads this month?”",
"options": {
"responseMode": "lastNode"
}
},
"id": "2de5c18f-62da-415c-a349-f01983763a93",
"name": "Bookkeeping chat",
"type": "@n8n/n8n-nodes-langchain.chatTrigger",
"typeVersion": 1.1,
"position": [
0,
0
],
"webhookId": "2de5c18f-62da-415c-a349-f01983763a93"
},
{
"parameters": {
"options": {
"systemMessage": "=You are a bookkeeping assistant for YOUR_BUSINESS. Currency: YOUR_CURRENCY. Today is {{ $now.toFormat('d LLLL yyyy') }}.\n\nTools:\n- \"Read ledger\": all transactions (date, type income/expense, amount, category, description).\n- \"Add transaction\": adds one income or expense. Categories: YOUR_CATEGORIES (e.g. Ads, Software, Salaries, Rent, Supplies, Sales, Other).\n- \"Email accountant\": drafts an email to our accountant (never sends).\n\nYou can: record transactions the user describes (\"paid 1,200 for Facebook ads yesterday\"), answer questions (\"how much did we spend on ads this month?\"), and prepare monthly summaries.\n\nRules:\n- Always read the ledger before answering questions about numbers. Do the arithmetic carefully and show the total and the period.\n- Confirm amount, date and category before adding a transaction.\n- You are not a tax advisor — for tax questions, suggest asking the accountant."
}
},
"id": "2de5c18f-62da-415c-a349-f01983763a93",
"name": "Bookkeeping agent",
"type": "@n8n/n8n-nodes-langchain.agent",
"typeVersion": 1.9,
"position": [
300,
0
]
},
{
"parameters": {
"model": {
"__rl": true,
"value": "claude-sonnet-5",
"mode": "list"
},
"options": {}
},
"id": "2de5c18f-62da-415c-a349-f01983763a93",
"name": "Claude model",
"type": "@n8n/n8n-nodes-langchain.lmChatAnthropic",
"typeVersion": 1.3,
"position": [
-40,
220
]
},
{
"parameters": {
"contextWindowLength": 10
},
"id": "2de5c18f-62da-415c-a349-f01983763a93",
"name": "Memory",
"type": "@n8n/n8n-nodes-langchain.memoryBufferWindow",
"typeVersion": 1.3,
"position": [
130,
220
]
},
{
"parameters": {
"descriptionType": "manual",
"toolDescription": "Returns all transactions: date, type (income/expense), amount, category, description.",
"operation": "read",
"documentId": {
"__rl": true,
"value": "YOUR_SPREADSHEET_ID",
"mode": "id"
},
"sheetName": {
"__rl": true,
"value": "Ledger",
"mode": "name"
},
"options": {}
},
"id": "2de5c18f-62da-415c-a349-f01983763a93",
"name": "Read ledger",
"type": "n8n-nodes-base.googleSheetsTool",
"typeVersion": 4.5,
"position": [
300,
220
]
},
{
"parameters": {
"descriptionType": "manual",
"toolDescription": "Adds one confirmed income or expense.",
"operation": "append",
"documentId": {
"__rl": true,
"value": "YOUR_SPREADSHEET_ID",
"mode": "id"
},
"sheetName": {
"__rl": true,
"value": "Ledger",
"mode": "name"
},
"columns": {
"mappingMode": "defineBelow",
"value": {
"date": "={{ $fromAI('date', \"Transaction date YYYY-MM-DD\", 'string') }}",
"type": "={{ $fromAI('type', \"income or expense\", 'string') }}",
"amount": "={{ $fromAI('amount', \"Amount as a number\", 'number') }}",
"category": "={{ $fromAI('category', \"One of the allowed categories\", 'string') }}",
"description": "={{ $fromAI('description', \"Short description\", 'string') }}"
}
},
"options": {}
},
"id": "2de5c18f-62da-415c-a349-f01983763a93",
"name": "Add transaction",
"type": "n8n-nodes-base.googleSheetsTool",
"typeVersion": 4.5,
"position": [
470,
220
]
},
{
"parameters": {
"descriptionType": "manual",
"toolDescription": "Drafts an email to the accountant (never sends it).",
"resource": "draft",
"operation": "create",
"subject": "={{ $fromAI('subject', \"Email subject\", 'string') }}",
"message": "={{ $fromAI('body', \"Email body, e.g. a monthly summary\", 'string') }}",
"options": {
"sendTo": "accountant@yourcompany.com"
}
},
"id": "2de5c18f-62da-415c-a349-f01983763a93",
"name": "Email accountant",
"type": "n8n-nodes-base.gmailTool",
"typeVersion": 2.1,
"position": [
640,
220
]
}
],
"connections": {
"Bookkeeping chat": {
"main": [
[
{
"node": "Bookkeeping agent",
"type": "main",
"index": 0
}
]
]
},
"Claude model": {
"ai_languageModel": [
[
{
"node": "Bookkeeping agent",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Memory": {
"ai_memory": [
[
{
"node": "Bookkeeping agent",
"type": "ai_memory",
"index": 0
}
]
]
},
"Read ledger": {
"ai_tool": [
[
{
"node": "Bookkeeping agent",
"type": "ai_tool",
"index": 0
}
]
]
},
"Add transaction": {
"ai_tool": [
[
{
"node": "Bookkeeping agent",
"type": "ai_tool",
"index": 0
}
]
]
},
"Email accountant": {
"ai_tool": [
[
{
"node": "Bookkeeping agent",
"type": "ai_tool",
"index": 0
}
]
]
}
},
"pinData": {},
"settings": {
"executionOrder": "v1"
},
"meta": {
"templateCredsSetupCompleted": false
},
"tags": []
}In Make, create a new scenario and add these modules in order:
- Webhooks › Custom webhook — called by your website's chat widget
- Make AI Agents › Run an agent — give it these tools: “Read ledger” (Google Sheets), “Add transaction” (Google Sheets), “Email accountant” (Gmail); paste the system prompt as the agent's instructions
- Every module run uses Make operations — the free plan includes a monthly quota.
- Module and step names can differ slightly as Make and Zapier update their apps.
In Zapier, create a new Zap and add these steps in order:
- Zapier Chatbots, or Webhooks by Zapier — Catch Hook from your chat widget
- Zapier Agents — create an agent (or AI by Zapier for a single step) — give it these tools: “Read ledger” (Google Sheets), “Add transaction” (Google Sheets), “Email accountant” (Gmail); paste the system prompt as the agent's instructions
- Zaps with more than two steps need a paid Zapier plan.
- Module and step names can differ slightly as Make and Zapier update their apps.
Tips
- Use a fixed list of categories so monthly reports stay clean.
Related flows
AI AgentsAdvanced
Customer support AI agent
A chat agent that answers from your FAQ, looks up orders, and hands tricky cases to a human.
AI model → Google Sheets → Gmail25h/mo
AI AgentsAdvanced
Unpaid invoices collections AI agent
Every morning, the agent reviews overdue invoices, drafts polite reminders in the right tone, and flags the ones that need a call.
AI model → Google Sheets → Gmail → Slack10h/mo
AI AgentsAdvanced
Inventory AI agent on Telegram
Ask "how many black abayas size M are left?" or say "received 40 of SKU-102"; the agent reads and updates your stock sheet.
Telegram → AI model → Google Sheets → Gmail10h/mo