This workflow follows the Agent → OpenAI Chat 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 →
{
"name": "Jour 6 - MCP guid\u00e9",
"nodes": [
{
"parameters": {
"content": "## Jour 6 - MCP guid\u00e9\n\nFlux minimal pour comparer une r\u00e9ponse na\u00efve et un acc\u00e8s outill\u00e9:\n1. D\u00e9clencheur manuel\n2. Question d'audit\n3. Agent de comparaison\n4. Tool MCP\n5. Sortie comparative\n\nObjectif: voir ce que le mod\u00e8le propose seul, puis ce que l'outil MCP permet de contr\u00f4ler.",
"height": 260,
"width": 640,
"color": 5
},
"id": "d6-note-overview",
"name": "Vue d'ensemble",
"type": "n8n-nodes-base.stickyNote",
"typeVersion": 1,
"position": [
0,
0
]
},
{
"parameters": {},
"id": "d6-trigger",
"name": "D\u00e9clencheur Manuel",
"type": "n8n-nodes-base.manualTrigger",
"typeVersion": 1,
"position": [
120,
380
]
},
{
"parameters": {
"mode": "manual",
"assignments": {
"assignments": [
{
"id": "d6-question",
"name": "chatInput",
"value": "Question m\u00e9tier: quelle r\u00e8gle d\u00e9crit l'\u00e9cart entre un montant URSSAF observ\u00e9 et le montant attendu base_urssaf * taux_urssaf ? Compare une r\u00e9ponse na\u00efve avec une r\u00e9ponse contr\u00f4l\u00e9e par MCP.",
"type": "string"
},
{
"id": "d6-naive-prompt",
"name": "naive_prompt",
"value": "R\u00e9ponds sans outil, comme si tu improvisais une hypoth\u00e8se rapide.",
"type": "string"
},
{
"id": "d6-tool-prompt",
"name": "tool_prompt",
"value": "Puis v\u00e9rifie la r\u00e9ponse avec le tool MCP search_documentary_sources et indique la diff\u00e9rence de gouvernance et de fiabilit\u00e9.",
"type": "string"
},
{
"id": "d6-question-tag",
"name": "question_tag",
"value": "audit_urssaf_compare_mcp",
"type": "string"
}
]
},
"options": {}
},
"id": "d6-question-node",
"name": "Question d'audit",
"type": "n8n-nodes-base.set",
"typeVersion": 3.4,
"position": [
380,
380
]
},
{
"parameters": {
"options": {
"systemMessage": "Tu es un agent de comparaison pour audit paie. Commence par une hypoth\u00e8se na\u00efve, puis utilise le tool MCP pour v\u00e9rifier la r\u00e8gle documentaire. Tu dois distinguer clairement: reponse_naive, reponse_controlee, comparaison, limites. N'invente pas de source et n'explique pas l'infrastructure MCP."
}
},
"id": "d6-agent-compare",
"name": "Agent de comparaison",
"type": "@n8n/n8n-nodes-langchain.agent",
"typeVersion": 1.9,
"position": [
660,
380
]
},
{
"parameters": {
"model": {
"__rl": true,
"value": "gpt-4.1-nano",
"mode": "id"
},
"options": {
"temperature": 0
}
},
"id": "d6-openai",
"name": "OpenAI Gateway",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"typeVersion": 1.2,
"position": [
660,
620
],
"credentials": {
"openAiApi": {
"name": "<your credential>"
}
}
},
{
"parameters": {
"sseEndpoint": "={{ $env.MCP_SSE_ENDPOINT || 'http://mcp-server:8000/sse' }}"
},
"id": "d6-mcp",
"name": "MCP outils guid\u00e9s",
"type": "@n8n/n8n-nodes-langchain.mcpClientTool",
"typeVersion": 1,
"position": [
660,
860
]
},
{
"parameters": {
"mode": "manual",
"assignments": {
"assignments": [
{
"id": "d6-output-status",
"name": "statut",
"value": "comparaison_achevee",
"type": "string"
},
{
"id": "d6-output-question",
"name": "question",
"value": "={{ $(\"Question d'audit\").item.json.chatInput }}",
"type": "string"
},
{
"id": "d6-output-report",
"name": "comparaison",
"value": "={{ $json.output }}",
"type": "string"
}
]
},
"options": {}
},
"id": "d6-comparison-output",
"name": "Sortie comparative",
"type": "n8n-nodes-base.set",
"typeVersion": 3.4,
"position": [
940,
380
]
}
],
"connections": {
"D\u00e9clencheur Manuel": {
"main": [
[
{
"node": "Question d'audit",
"type": "main",
"index": 0
}
]
]
},
"Question d'audit": {
"main": [
[
{
"node": "Agent de comparaison",
"type": "main",
"index": 0
}
]
]
},
"Agent de comparaison": {
"main": [
[
{
"node": "Sortie comparative",
"type": "main",
"index": 0
}
]
]
},
"OpenAI Gateway": {
"ai_languageModel": [
[
{
"node": "Agent de comparaison",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"MCP outils guid\u00e9s": {
"ai_tool": [
[
{
"node": "Agent de comparaison",
"type": "ai_tool",
"index": 0
}
]
]
}
},
"settings": {
"executionOrder": "v1"
},
"meta": {
"description": "Jour 6 - MCP guid\u00e9: une question m\u00e9tier, une comparaison entre r\u00e9ponse na\u00efve et r\u00e9ponse outill\u00e9e, et un seul client MCP sans debug d'infrastructure."
}
}
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.
openAiApi
For the full experience including quality scoring and batch install features for each workflow upgrade to Pro
About this workflow
Jour 6 - MCP guidé. Uses agent, lmChatOpenAi, mcpClientTool. Event-driven trigger; 7 nodes.
Source: https://github.com/ssime-git/deloitte-ia-nocode/blob/a1a6947327f3668e93488e1724ce3147dcedf09b/n8n/workflows/day6_mcp_guided.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.
This assistant is designed for people who don't have time to write and send emails to suppliers. With just one request, it drafts and sends clear, professional messages automatically.
Sales reps, agencies, and growth teams who want to turn basic company info into qualified leads with automated research . Perfect for B2B prospecting.
Stop switching between Slack and your analytics dashboards. Mention the bot in any Slack channel, ask a business question, and get an AI-powered answer from Databox in seconds - without leaving Slack.
search. Uses executeWorkflowTrigger, agent, lmChatOpenAi, mcpClientTool. Event-driven trigger; 4 nodes.
music. Uses executeWorkflowTrigger, agent, lmChatOpenAi, mcpClientTool. Event-driven trigger; 4 nodes.