Yaron Been This workflow corresponds to n8n.io template #6081 — we link there as the canonical source.
This workflow follows the Agent → Form Trigger 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": "HyjRDR7O78FdTjJM",
"meta": {
"templateCredsSetupCompleted": true
},
"name": "6. Generate and Auto-Evaluate Facebook Ad Headlines using GPT-4o-mini",
"tags": [],
"nodes": [
{
"id": "53bc6494-458c-4dcb-a66c-092fb30bf6b7",
"name": "Send a message",
"type": "n8n-nodes-base.gmail",
"position": [
2784,
144
],
"parameters": {
"message": "={{ $('Agent_HeadlineWriter').item.json.output }}",
"options": {}
},
"typeVersion": 2.1
},
{
"id": "ca661097-1b4e-4dc6-b22c-2ff220106928",
"name": "FormTrigger_CopywritingBrief",
"type": "n8n-nodes-base.formTrigger",
"position": [
256,
400
],
"parameters": {
"options": {},
"formTitle": "Copywriting Agent",
"formFields": {
"values": [
{
"fieldLabel": "What is your product about?",
"requiredField": true
}
]
}
},
"typeVersion": 2.2
},
{
"id": "340fb887-e9e7-45d6-8294-09337c5b20fc",
"name": "Set_PromptForHeadline",
"type": "n8n-nodes-base.set",
"position": [
464,
432
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "e1ad4483-fb6d-4a4a-8589-615fbe8530c9",
"name": "Prompt For Agent",
"type": "string",
"value": "Write a facebook ad headline for this product:"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "08915a22-5c32-4680-a299-9588fffc08b2",
"name": "LLM_HeadlineWriterModel",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
784,
608
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "gpt-4o-mini"
},
"options": {}
},
"typeVersion": 1.2
},
{
"id": "a831df58-17e1-4e56-863b-e5918b1a4e80",
"name": "Agent_HeadlineWriter",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
816,
384
],
"parameters": {
"text": "={{ $json['Prompt For Agent'] }}:{{ $('FormTrigger_CopywritingBrief').item.json['What is your product about?'] }}\n\n",
"options": {},
"promptType": "define"
},
"typeVersion": 1.8
},
{
"id": "fd5cdd53-4738-4023-ac23-f8ada2c176f9",
"name": "LLM_EvalCriteriaModel",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
1216,
576
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "gpt-4o-mini"
},
"options": {}
},
"typeVersion": 1.2
},
{
"id": "d9e284c1-92d1-43c9-9307-a4b3a2d3479f",
"name": "LLM_HeadlineEvaluatorModel",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
1616,
592
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "gpt-4o-mini"
},
"options": {}
},
"typeVersion": 1.2
},
{
"id": "2484f5c0-6d3b-4a81-a82e-bb9619bc27c6",
"name": "LLM_BottomLineModel",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
2128,
624
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "gpt-4o-mini"
},
"options": {}
},
"typeVersion": 1.2
},
{
"id": "15d3350a-4fe1-49dd-8a2a-3f03350c18a2",
"name": "Agent_IterationDecision",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
2144,
384
],
"parameters": {
"text": "=Based on the bottom line:{{ $json.output }}\n\n\nDecide if we need another iteration.\nReturn NO if we dont,\notherwise return YES and add your feedback.",
"options": {},
"promptType": "define"
},
"typeVersion": 1.8
},
{
"id": "086982ae-6aef-43c8-8cd1-228a2fcaebe0",
"name": "If_NeedMoreIterations",
"type": "n8n-nodes-base.if",
"position": [
2576,
224
],
"parameters": {
"options": {},
"conditions": {
"options": {
"version": 2,
"leftValue": "",
"caseSensitive": true,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "7f7d7b1c-e8ae-4a87-a8fe-43e32f270b64",
"operator": {
"name": "filter.operator.equals",
"type": "string",
"operation": "equals"
},
"leftValue": "={{ $json.output }}",
"rightValue": "NO"
}
]
}
},
"typeVersion": 2.2
},
{
"id": "02058dda-acda-4d7d-bf31-062598f91683",
"name": "Agent_HeadlineEvaluator",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
1712,
384
],
"parameters": {
"text": "=Your goal is to asses the output of an AI agent.\nIt was supposed to:\n{{ $('Set_PromptForHeadline').item.json['Prompt For Agent'] }}\n\nThe answer it gave is:{{ $('Agent_HeadlineWriter').item.json.output }}\n\nProvide your response using these parameters:{{ $json.output }}. make sure you calculate your average and give a bottom line.",
"options": {},
"promptType": "define"
},
"typeVersion": 1.8
},
{
"id": "14734fe9-5368-40b6-b386-df80aa281507",
"name": "Agent_EvalCriteriaBuilder",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
1264,
384
],
"parameters": {
"text": "=Your goal is to define parametes for assesing the output of this prompt:\n{{ $('Set_PromptForHeadline').item.json['Prompt For Agent'] }}\n\nWhich parameters would you suggest? give me 5 parameters including the scale. All should be in scale of 1-10 (10 will be the most positive)",
"options": {},
"promptType": "define"
},
"typeVersion": 1.8
},
{
"id": "ff7e0559-fc28-4be8-bf1b-492e4cbcb3f9",
"name": "Sticky Note9",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1216,
-272
],
"parameters": {
"color": 4,
"width": 1300,
"height": 320,
"content": "=======================================\n WORKFLOW ASSISTANCE\n=======================================\nFor any questions or support, please contact:\n Yaron@nofluff.online\n\nExplore more tips and tutorials here:\n - YouTube: https://www.youtube.com/@YaronBeen/videos\n - LinkedIn: https://www.linkedin.com/in/yaronbeen/\n=======================================\n"
},
"typeVersion": 1
},
{
"id": "d3cbada6-0cbe-42cb-b7a9-600bd1c1a567",
"name": "Sticky Note4",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1216,
96
],
"parameters": {
"color": 4,
"width": 1289,
"height": 2214,
"content": "# Generate and Auto-Evaluate Facebook Ad Headlines using GPT-4o-mini\n\nBuilt with **n8n + OpenAI**\n\nThis workflow captures a product description, generates ad headlines, evaluates them with custom criteria, decides whether another draft is needed, and finally sends the best version via Gmail.\n\n---\n\n## \u26a1 Section 1: Capture the Brief & Build the Prompt\n\n* **\ud83d\udcdd FormTrigger\\_CopywritingBrief** \u2192 A simple form asks: *\u201cWhat is your product about?\u201d*\n* **\u2699\ufe0f Set\\_PromptForHeadline** \u2192 Prepares the input by appending the instruction:\n *\u201cWrite a Facebook ad headline for this product:\u201d*\n\n**Benefit:** Ensures consistent, structured prompts so the AI receives clear context every time.\n\n---\n\n## \u270d\ufe0f Section 2: Draft the Headline\n\n* **\ud83d\udcac LLM\\_HeadlineWriterModel** \u2192 GPT-4o-mini model provides the intelligence.\n* **\u270d\ufe0f Agent\\_HeadlineWriter** \u2192 Generates a first-pass Facebook ad headline.\n\n**Benefit:** Produces creative copy instantly without waiting on a human writer.\n\n---\n\n## \ud83d\udccb Section 3: Define Scoring Criteria\n\n* **\ud83d\udcac LLM\\_EvalCriteriaModel** \u2192 Calls GPT-4o-mini again.\n* **\ud83d\udcd1 Agent\\_EvalCriteriaBuilder** \u2192 Suggests 5 scoring parameters (scale 1-10).\n Example: Clarity, Relevance, Hook Strength, Brand Voice, Scroll-Stoppage.\n\n**Benefit:** Builds an objective, repeatable evaluation rubric automatically.\n\n---\n\n## \ud83d\udd0d Section 4: Evaluate the Headline\n\n* **\ud83d\udcac LLM\\_HeadlineEvaluatorModel** \u2192 Supplies reasoning power.\n* **\ud83d\udd0d Agent\\_HeadlineEvaluator** \u2192 Applies the 5 criteria to the generated headline and outputs:\n\n * JSON with scores per parameter\n * An average score\n * A plain-language bottom-line\n\n**Benefit:** Turns subjective copy quality into measurable numbers.\n\n---\n\n## \ud83d\udd04 Section 5: Decide & Iterate (if needed)\n\n* **\ud83d\udcac LLM\\_BottomLineModel** \u2192 Interprets the evaluation results.\n* **\ud83e\udd14 Agent\\_IterationDecision** \u2192 Decides:\n\n * Return **NO** \u2192 headline is acceptable.\n * Return **YES** + feedback \u2192 headline should be rewritten.\n* **\ud83d\udd00 If\\_NeedMoreIterations** \u2192 Branches:\n\n * If **NO** \u2192 continue workflow.\n * If **YES** \u2192 (loop wiring possible) headline can be regenerated with feedback.\n\n**Benefit:** Keeps iterating until the AI headline meets your standards.\n\n---\n\n## \ud83d\udce9 Section 6: Deliver the Result\n\n* **\ud83d\udce7 Send a message (Gmail node)** \u2192 Sends the accepted headline via email.\n\n**Benefit:** Automates delivery of the polished, AI-approved headline to your inbox or team.\n\n---\n\n## \ud83d\udcca Workflow Overview\n\n| Section | Purpose | Key Nodes | Benefit |\n| -------------------- | ---------------------------------- | ----------------------------------------------------- | ------------------------------ |\n| \u26a1 Capture Brief | Collect product info & prep prompt | FormTrigger, Set | Structured AI input |\n| \u270d\ufe0f Draft Headline | Generate first headline | LLM\\_HeadlineWriterModel, Agent\\_HeadlineWriter | Instant creative draft |\n| \ud83d\udccb Define Criteria | Build scoring rubric | LLM\\_EvalCriteriaModel, Agent\\_EvalCriteriaBuilder | Objective evaluation |\n| \ud83d\udd0d Evaluate Headline | Score headline & summarize | LLM\\_HeadlineEvaluatorModel, Agent\\_HeadlineEvaluator | Transparent quality check |\n| \ud83d\udd04 Decide & Iterate | Accept or refine headline | LLM\\_BottomLineModel, Agent\\_IterationDecision, If | Only good results move forward |\n| \ud83d\udce9 Deliver Result | Share the final copy | Gmail | Automates delivery |\n\n---\n\n## \u2705 Final Benefits\n\n* \ud83d\ude80 One-click workflow: from product description to tested headline.\n* \ud83d\udcca Automatic rubric: objective scoring each time.\n* \ud83d\udd04 Self-improving: poor headlines can auto-iterate with feedback.\n* \ud83d\udce7 Direct integration: approved headlines land in Gmail instantly.\n* \ud83e\udde9 Fully modular: easy to extend with Google Sheets, Slack, or CRM nodes.\n\n---\n\n"
},
"typeVersion": 1
}
],
"active": false,
"settings": {
"executionOrder": "v1"
},
"versionId": "5f62337f-55ba-47a5-b3c4-0fe66cf02527",
"connections": {
"LLM_BottomLineModel": {
"ai_languageModel": [
[
{
"node": "Agent_IterationDecision",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Agent_HeadlineWriter": {
"main": [
[
{
"node": "Agent_EvalCriteriaBuilder",
"type": "main",
"index": 0
}
]
]
},
"If_NeedMoreIterations": {
"main": [
[
{
"node": "Send a message",
"type": "main",
"index": 0
}
]
]
},
"LLM_EvalCriteriaModel": {
"ai_languageModel": [
[
{
"node": "Agent_EvalCriteriaBuilder",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Set_PromptForHeadline": {
"main": [
[
{
"node": "Agent_HeadlineWriter",
"type": "main",
"index": 0
}
]
]
},
"Agent_HeadlineEvaluator": {
"main": [
[
{
"node": "Agent_IterationDecision",
"type": "main",
"index": 0
}
]
]
},
"Agent_IterationDecision": {
"main": [
[
{
"node": "If_NeedMoreIterations",
"type": "main",
"index": 0
}
]
]
},
"LLM_HeadlineWriterModel": {
"ai_languageModel": [
[
{
"node": "Agent_HeadlineWriter",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"Agent_EvalCriteriaBuilder": {
"main": [
[
{
"node": "Agent_HeadlineEvaluator",
"type": "main",
"index": 0
}
]
]
},
"LLM_HeadlineEvaluatorModel": {
"ai_languageModel": [
[
{
"node": "Agent_HeadlineEvaluator",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"FormTrigger_CopywritingBrief": {
"main": [
[
{
"node": "Set_PromptForHeadline",
"type": "main",
"index": 0
}
]
]
}
}
}
Pro
For the full experience including quality scoring and batch install features for each workflow upgrade to Pro
About this workflow
Built with n8n + OpenAI
Source: https://n8n.io/workflows/6081/ — 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.
This workflow automates end-to-end contract and invoice management using AI intelligence. It processes proposals through intelligent contract generation, approval workflows, and automated invoicing. O
Form Trigger, Data Table, Agent +4
This n8n workflow template automates your lead generation and follow-up process using AI. It captures leads through a form, enriches them with company data, classifies them into different categories,
Gmail, Output Parser Structured, Form Trigger +5
Automates SaaS operations by consolidating user management, AI-driven support triage, analytics, and billing into one unified system. User signups flow through registration, support requests route via
Form Trigger, Data Table, Agent +7
The workflow runs every hour with a randomized delay of 5–20 minutes to help distribute load. It records the exact date and time a lead is emailed so you can track outreach. Follow-ups are automatical
Google Sheets, Agent, OpenAI Chat +5
This n8n workflow automates turning short user ideas into production-ready real-estate marketing assets (photorealistic images and optional 360° videos). A form submission seeds a prompt board → an LL
Form Trigger, Google Sheets, Agent +6