Agent Marketing - Tinyllama (pret a Importer)

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": "Agent Marketing - TinyLlama (pret a importer)",
  "nodes": [
    {
      "parameters": {
        "options": {
          "responseMode": "whenLastNodeFinishes"
        }
      },
      "id": "trigger-001",
      "name": "Chat Trigger",
      "type": "@n8n/n8n-nodes-langchain.chatTrigger",
      "position": [
        220,
        380
      ],
      "typeVersion": 1
    },
    {
      "parameters": {
        "jsCode": "const chatInput = $input.first().json.chatInput || '';\nconst urlMatch = chatInput.match(/https?:\\/\\/[^\\s]+/);\nconst url = urlMatch ? urlMatch[0].trim() : 'https://example.com';\nreturn [{ json: { chatInput, url } }];"
      },
      "id": "code-extract",
      "name": "Extract URL",
      "type": "n8n-nodes-base.code",
      "position": [
        440,
        380
      ],
      "typeVersion": 2
    },
    {
      "parameters": {
        "url": "={{ $json.url }}",
        "options": {
          "allowUnauthorizedCerts": true,
          "response": {
            "response": {
              "responseFormat": "text"
            }
          }
        }
      },
      "id": "http-fetch",
      "name": "Fetch Page",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        660,
        380
      ],
      "typeVersion": 4.2
    },
    {
      "parameters": {
        "jsCode": "const chatInput = $('Extract URL').first().json.chatInput;\nconst raw = $input.first().json;\nlet pageContent = raw.body ?? raw.data ?? raw.text ?? JSON.stringify(raw);\nif (typeof pageContent === 'string') {\n  pageContent = pageContent.replace(/<script[^>]*>[\\s\\S]*?<\\/script>/gi, '').replace(/<style[^>]*>[\\s\\S]*?<\\/style>/gi, '').replace(/<[^>]+>/g, ' ').replace(/\\s+/g, ' ').trim();\n  if (pageContent.length > 900) pageContent = pageContent.slice(0, 900) + '...';\n}\nreturn [{ json: { chatInput, pageContent } }];"
      },
      "id": "code-prepare",
      "name": "Prepare for LLM",
      "type": "n8n-nodes-base.code",
      "position": [
        880,
        380
      ],
      "typeVersion": 2
    },
    {
      "parameters": {
        "jsCode": "const { chatInput, pageContent } = $input.first().json;\nconst trimmedInput = (chatInput || '').trim();\nconst userMessage = `Question utilisateur : ${trimmedInput}\\n\\nContenu du site : ${pageContent}\\n\\nR\u00e9ponds en fran\u00e7ais avec des puces courtes.`;\nconst body = {\n  model: 'tinyllama',\n  stream: false,\n  messages: [\n    { role: 'system', content: 'Tu es un consultant marketing. R\u00e9ponds bri\u00e8vement, sans r\u00e9p\u00e9ter la question.' },\n    { role: 'user', content: userMessage }\n  ]\n};\nconst response = await this.helpers.httpRequest({\n  method: 'POST',\n  url: 'http://host.docker.internal:11434/api/chat',\n  body,\n  json: true,\n  skipSslCertificateValidation: true\n});\nconst content = response.message?.content ?? response.body?.message?.content ?? response.content;\nconst text = (typeof content === 'string' ? content : (response.message || response.body?.message || 'Pas de r\u00e9ponse.'));\nlet out = typeof text === 'string' ? text : (text?.content ?? JSON.stringify(text));\nif (typeof out === 'string') {\n  out = out.replace(/^Question utilisateur\\s*:\\s*/i, '').trim();\n  if (/^Tu es un consultant/i.test(out)) out = out.replace(/^Tu es un consultant[^\\n]*\\.?\\s*\\n?/gi, '').trim();\n}\nif (typeof out === 'string' && out.length < 30) out = 'TinyLlama n\\'a pas g\u00e9n\u00e9r\u00e9 de r\u00e9ponse exploitable. R\u00e9essaie avec une demande plus courte.';\nreturn [{ json: { text: out, output: out } }];"
      },
      "id": "code-ollama",
      "name": "Call Ollama",
      "type": "n8n-nodes-base.code",
      "position": [
        1100,
        380
      ],
      "typeVersion": 2
    },
    {
      "parameters": {
        "content": "## Apr\u00e8s import :\n\n1. Aucune credential \u00e0 configurer (Ollama est appel\u00e9 en direct dans le n\u0153ud Call Ollama).\n\n2. Save workflow (Ctrl+S) > Activer (interrupteur en haut) > Chat en bas pour tester.\n\n3. Exemple : \u00ab Analyse le site https://example.com et propose une strat\u00e9gie de contenu marketing \u00bb\n\nOllama doit tourner sur ta machine avec : ollama pull tinyllama",
        "height": 240,
        "width": 320,
        "color": 4
      },
      "id": "note-001",
      "name": "Sticky Note",
      "type": "n8n-nodes-base.stickyNote",
      "position": [
        180,
        120
      ],
      "typeVersion": 1
    }
  ],
  "connections": {
    "Chat Trigger": {
      "main": [
        [
          {
            "node": "Extract URL",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Extract URL": {
      "main": [
        [
          {
            "node": "Fetch Page",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Fetch Page": {
      "main": [
        [
          {
            "node": "Prepare for LLM",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Prepare for LLM": {
      "main": [
        [
          {
            "node": "Call Ollama",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  },
  "settings": {
    "executionOrder": "v1"
  },
  "meta": {
    "templateCredsSetupCompleted": false
  }
}

About this workflow

Agent Marketing - TinyLlama (pret a importer). Uses chatTrigger, httpRequest. Chat trigger; 6 nodes.

Source: https://github.com/CHANG-Toma/Agent-IA/blob/0a54ca4159d4650126d2d21252909f02377aa7a3/03-Workflows/agent-marketing-tinyllama-import.json — original creator credit. Request a take-down →

More AI & RAG workflows → · Browse all categories →

Related workflows

Workflows that share integrations, category, or trigger type with this one. All free to copy and import.