Hourly RSS Feed Processor Automation

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": "Kairos - RSS Processor v3",
  "nodes": [
    {
      "parameters": {
        "rule": {
          "interval": [
            {
              "field": "hours",
              "hoursInterval": 1
            }
          ]
        }
      },
      "id": "schedule",
      "name": "Toutes les heures",
      "type": "n8n-nodes-base.scheduleTrigger",
      "typeVersion": 1.2,
      "position": [
        -200,
        300
      ]
    },
    {
      "parameters": {
        "httpMethod": "POST",
        "path": "rss-process",
        "options": {}
      },
      "id": "webhook",
      "name": "Webhook",
      "type": "n8n-nodes-base.webhook",
      "typeVersion": 2,
      "position": [
        -200,
        500
      ]
    },
    {
      "parameters": {
        "method": "POST",
        "url": "http://kairos-rest:3000/rpc/get_active_topics_with_feeds",
        "sendHeaders": true,
        "headerParameters": {
          "parameters": [
            {
              "name": "apikey",
              "value": "CHANGE_ME_SERVICE_ROLE_KEY"
            },
            {
              "name": "Authorization",
              "value": "Bearer CHANGE_ME_SERVICE_ROLE_KEY"
            },
            {
              "name": "Content-Type",
              "value": "application/json"
            }
          ]
        },
        "options": {}
      },
      "id": "get-topics",
      "name": "Get Topics",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.2,
      "position": [
        50,
        400
      ]
    },
    {
      "parameters": {
        "conditions": {
          "options": {
            "caseSensitive": true,
            "leftValue": "",
            "typeValidation": "strict"
          },
          "conditions": [
            {
              "id": "topics-exist",
              "leftValue": "={{ $json }}",
              "rightValue": "",
              "operator": {
                "type": "array",
                "operation": "notEmpty"
              }
            }
          ],
          "combinator": "and"
        },
        "options": {}
      },
      "id": "if-topics",
      "name": "Topics existent?",
      "type": "n8n-nodes-base.if",
      "typeVersion": 2,
      "position": [
        270,
        400
      ]
    },
    {
      "parameters": {
        "jsCode": "// Extraire les feeds RSS de chaque topic\nconst items = $input.all();\nconst results = [];\n\nfor (const item of items) {\n  const topic = item.json;\n  if (!topic || !topic.rss_feeds) continue;\n\n  const feeds = topic.rss_feeds || [];\n  \n  for (const feedUrl of feeds) {\n    results.push({\n      json: {\n        topic_id: topic.topic_id,\n        topic_name: topic.topic_name,\n        user_id: topic.user_id,\n        keywords_fr: topic.keywords_fr || [],\n        keywords_en: topic.keywords_en || [],\n        feed_url: feedUrl\n      }\n    });\n  }\n}\n\nif (results.length === 0) {\n  return [{ json: { error: 'No feeds found' } }];\n}\nreturn results;"
      },
      "id": "extract-feeds",
      "name": "Extract Feeds",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        490,
        300
      ]
    },
    {
      "parameters": {
        "url": "={{ $json.feed_url }}",
        "options": {}
      },
      "id": "fetch-rss",
      "name": "Fetch RSS",
      "type": "n8n-nodes-base.rssFeedRead",
      "typeVersion": 1.1,
      "position": [
        710,
        300
      ],
      "onError": "continueRegularOutput",
      "continueOnFail": true
    },
    {
      "parameters": {
        "jsCode": "// Parser les articles RSS et preparer les donnees\nconst items = $input.all();\nconst feedInfo = $('Extract Feeds').first().json;\nconst results = [];\n\nfor (const item of items) {\n  if (!item.json.link || !item.json.title) continue;\n\n  const title = item.json.title.trim();\n  let content = item.json.content || item.json.description || '';\n  content = content.replace(/<[^>]*>/g, ' ')\n    .replace(/&nbsp;/g, ' ')\n    .replace(/&amp;/g, '&')\n    .replace(/&lt;/g, '<')\n    .replace(/&gt;/g, '>')\n    .replace(/&quot;/g, '\"')\n    .replace(/\\s+/g, ' ')\n    .trim()\n    .substring(0, 3000);\n\n  results.push({\n    json: {\n      topic_id: feedInfo.topic_id,\n      topic_name: feedInfo.topic_name,\n      user_id: feedInfo.user_id,\n      keywords_fr: feedInfo.keywords_fr,\n      keywords_en: feedInfo.keywords_en,\n      feed_url: feedInfo.feed_url,\n      title: title,\n      url: item.json.link,\n      content_rss: content,\n      author: item.json.creator || item.json.author || null,\n      published_at: item.json.pubDate || item.json.isoDate || new Date().toISOString(),\n      image_url: item.json.enclosure?.url || null,\n      source: feedInfo.feed_url.match(/https?:\\/\\/([^/]+)/)?.[1] || 'Unknown'\n    }\n  });\n}\n\n// Limiter a 10 articles par feed\nreturn results.slice(0, 10);"
      },
      "id": "parse-articles",
      "name": "Parse Articles",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        930,
        300
      ]
    },
    {
      "parameters": {
        "batchSize": 1,
        "options": {}
      },
      "id": "loop-articles",
      "name": "Loop Articles",
      "type": "n8n-nodes-base.splitInBatches",
      "typeVersion": 3,
      "position": [
        1150,
        300
      ]
    },
    {
      "parameters": {
        "method": "GET",
        "url": "=http://kairos-rest:3000/articles?topic_id=eq.{{ $json.topic_id }}&url=eq.{{ encodeURIComponent($json.url) }}&select=id&limit=1",
        "sendHeaders": true,
        "headerParameters": {
          "parameters": [
            {
              "name": "apikey",
              "value": "CHANGE_ME_SERVICE_ROLE_KEY"
            },
            {
              "name": "Authorization",
              "value": "Bearer CHANGE_ME_SERVICE_ROLE_KEY"
            }
          ]
        },
        "options": {}
      },
      "id": "check-duplicate",
      "name": "Check Duplicate",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.2,
      "position": [
        1370,
        300
      ],
      "onError": "continueRegularOutput"
    },
    {
      "parameters": {
        "conditions": {
          "options": {
            "caseSensitive": true,
            "leftValue": "",
            "typeValidation": "loose"
          },
          "conditions": [
            {
              "id": "not-exists",
              "leftValue": "={{ $json.id }}",
              "rightValue": "",
              "operator": {
                "type": "string",
                "operation": "notExists"
              }
            }
          ],
          "combinator": "and"
        },
        "options": {}
      },
      "id": "if-new",
      "name": "Nouvel article?",
      "type": "n8n-nodes-base.if",
      "typeVersion": 2,
      "position": [
        1590,
        300
      ]
    },
    {
      "parameters": {
        "url": "=https://r.jina.ai/{{ $('Loop Articles').item.json.url }}",
        "options": {
          "response": {
            "response": {
              "responseFormat": "text"
            }
          },
          "timeout": 30000
        }
      },
      "id": "jina-extract",
      "name": "Jina Extract",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.2,
      "position": [
        1810,
        200
      ],
      "retryOnFail": true,
      "maxTries": 2,
      "waitBetweenTries": 3000,
      "onError": "continueRegularOutput"
    },
    {
      "parameters": {
        "jsCode": "// Combiner les donnees de l'article avec le contenu Jina\nconst articleData = $('Loop Articles').first().json;\nconst jinaResponse = $input.first().json;\n\n// Utiliser le contenu Jina si disponible, sinon le contenu RSS\nlet content = jinaResponse.data || articleData.content_rss || '';\ncontent = content.substring(0, 8000); // Limiter la taille\n\n// Combiner les mots-cles\nconst keywords = [\n  ...(articleData.keywords_fr || []),\n  ...(articleData.keywords_en || [])\n].join(', ');\n\nreturn [{\n  json: {\n    topic_id: articleData.topic_id,\n    topic_name: articleData.topic_name,\n    user_id: articleData.user_id,\n    keywords: keywords,\n    feed_url: articleData.feed_url,\n    title: articleData.title,\n    url: articleData.url,\n    content: content,\n    author: articleData.author,\n    published_at: articleData.published_at,\n    image_url: articleData.image_url,\n    source: articleData.source\n  }\n}];"
      },
      "id": "merge-content",
      "name": "Merge Content",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        2030,
        200
      ]
    },
    {
      "parameters": {
        "method": "POST",
        "url": "http://kairos-ollama:11434/api/generate",
        "sendBody": true,
        "specifyBody": "json",
        "jsonBody": "={\n  \"model\": \"gemma3:4b\",\n  \"prompt\": \"Resume cet article en 2-3 phrases concises en francais. Concentre-toi sur les informations cles et les points importants. Ne commence pas par 'Cet article' ou 'L'article'. Reponds directement avec le resume.\\n\\nTitre: {{ $json.title }}\\n\\nContenu: {{ $json.content.substring(0, 4000).replace(/[\\n\\r]+/g, ' ').replace(/\\\\/g, '\\\\\\\\').replace(/\"/g, '\\\\\"') }}\",\n  \"system\": \"Tu es un assistant specialise dans le resume d'articles de veille technologique. Tu reponds toujours en francais, de maniere concise et factuelle.\",\n  \"stream\": false,\n  \"options\": {\n    \"temperature\": 0.3,\n    \"num_predict\": 200\n  }\n}",
        "options": {
          "timeout": 120000
        }
      },
      "id": "ollama-summary",
      "name": "IA - Resume",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.2,
      "position": [
        2250,
        200
      ],
      "retryOnFail": true,
      "maxTries": 2,
      "waitBetweenTries": 2000,
      "onError": "continueRegularOutput"
    },
    {
      "parameters": {
        "jsCode": "// Extraire le resume et preparer pour la pertinence\nconst articleData = $('Merge Content').first().json;\nconst summaryResponse = $input.first().json;\n\nlet summary = summaryResponse?.response || '';\nsummary = summary.trim().substring(0, 1000);\n\nreturn [{\n  json: {\n    topic_id: articleData.topic_id,\n    topic_name: articleData.topic_name,\n    user_id: articleData.user_id,\n    keywords: articleData.keywords,\n    feed_url: articleData.feed_url,\n    title: articleData.title,\n    url: articleData.url,\n    content: articleData.content,\n    author: articleData.author,\n    published_at: articleData.published_at,\n    image_url: articleData.image_url,\n    source: articleData.source,\n    summary: summary\n  }\n}];"
      },
      "id": "extract-summary",
      "name": "Extract Summary",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        2470,
        200
      ]
    },
    {
      "parameters": {
        "method": "POST",
        "url": "http://kairos-ollama:11434/api/generate",
        "sendBody": true,
        "specifyBody": "json",
        "jsonBody": "={\n  \"model\": \"gemma3:4b\",\n  \"prompt\": \"Tu dois noter la PERTINENCE de cet article par rapport au sujet de veille.\\n\\nSUJET DE VEILLE: {{ $json.topic_name }}\\nMOTS-CLES DU SUJET: {{ $json.keywords }}\\n\\nARTICLE A EVALUER:\\nTitre: {{ $json.title }}\\nContenu: {{ $json.summary.replace(/[\\n\\r]+/g, ' ').replace(/\\\\/g, '\\\\\\\\').replace(/\"/g, '\\\\\"') }}\\n\\nINSTRUCTIONS:\\nDonne une note PRECISE entre 0 et 100 (pas une note sur 5 ou 10).\\n\\nExemples de notes attendues:\\n- Un article 100% sur le sujet = 92 ou 87 ou 95\\n- Un article partiellement lie = 58 ou 63 ou 71\\n- Un article peu lie = 34 ou 41 ou 28\\n- Un article sans rapport = 12 ou 8 ou 15\\n\\nATTENTION: Ne reponds PAS par 1, 2, 3, 4 ou 5. Reponds par un nombre entre 0 et 100.\\n\\nTA NOTE (un seul nombre entre 0 et 100):\",\n  \"system\": \"Tu es un evaluateur. Tu donnes une note de pertinence entre 0 et 100. Tu reponds UNIQUEMENT par un nombre entier. Exemples de reponses valides: 73, 45, 88, 91, 34, 67. Exemples de reponses INVALIDES: 4, 5, 3/5, quatre.\",\n  \"stream\": false,\n  \"options\": {\n    \"temperature\": 0.3,\n    \"num_predict\": 8\n  }\n}",
        "options": {
          "timeout": 60000
        }
      },
      "id": "ollama-relevance",
      "name": "IA - Pertinence",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.2,
      "position": [
        2690,
        200
      ],
      "retryOnFail": true,
      "maxTries": 2,
      "waitBetweenTries": 2000,
      "onError": "continueRegularOutput"
    },
    {
      "parameters": {
        "jsCode": "// Extraire le score de pertinence\nconst articleData = $('Extract Summary').first().json;\nconst relevanceResponse = $input.first().json;\n\nlet relevanceScore = 50; // Valeur par defaut\nif (relevanceResponse?.response) {\n  const match = relevanceResponse.response.match(/\\d+/);\n  if (match) {\n    relevanceScore = Math.min(100, Math.max(0, parseInt(match[0])));\n  }\n}\n\nreturn [{\n  json: {\n    topic_id: articleData.topic_id,\n    topic_name: articleData.topic_name,\n    user_id: articleData.user_id,\n    keywords: articleData.keywords,\n    feed_url: articleData.feed_url,\n    title: articleData.title,\n    url: articleData.url,\n    content: articleData.content,\n    author: articleData.author,\n    published_at: articleData.published_at,\n    image_url: articleData.image_url,\n    source: articleData.source,\n    summary: articleData.summary,\n    relevance_score: relevanceScore\n  }\n}];"
      },
      "id": "extract-relevance",
      "name": "Extract Relevance",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        2910,
        200
      ]
    },
    {
      "parameters": {
        "method": "POST",
        "url": "http://kairos-ollama:11434/api/generate",
        "sendBody": true,
        "specifyBody": "json",
        "jsonBody": "={\n  \"model\": \"gemma3:4b\",\n  \"prompt\": \"Genere 3 a 5 tags pertinents pour cet article. Les tags doivent etre en minuscules, en francais, separes par des virgules, sans espaces dans les tags (utilise des tirets).\\n\\nExemples de tags: intelligence-artificielle, securite, cloud, startup, open-source, machine-learning, api, database, devops, frontend, backend, cybersecurite, automatisation, data-science, blockchain, fintech, saas, iot, robotique\\n\\nReponds UNIQUEMENT avec les tags separes par des virgules, rien d'autre.\\n\\nTitre: {{ $json.title }}\\nResume: {{ $json.summary.replace(/[\\n\\r]+/g, ' ').replace(/\\\\/g, '\\\\\\\\').replace(/\"/g, '\\\\\"') }}\",\n  \"system\": \"Tu es un classificateur d'articles. Tu generes des tags pertinents en minuscules, separes par des virgules.\",\n  \"stream\": false,\n  \"options\": {\n    \"temperature\": 0.3,\n    \"num_predict\": 50\n  }\n}",
        "options": {
          "timeout": 60000
        }
      },
      "id": "ollama-tags",
      "name": "IA - Tags",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.2,
      "position": [
        3130,
        200
      ],
      "retryOnFail": true,
      "maxTries": 2,
      "waitBetweenTries": 2000,
      "onError": "continueRegularOutput"
    },
    {
      "parameters": {
        "jsCode": "// Extraire les tags et preparer les donnees finales\nconst articleData = $('Extract Relevance').first().json;\nconst tagsResponse = $input.first().json;\n\n// Parser les tags\nlet tags = [];\nif (tagsResponse?.response) {\n  tags = tagsResponse.response\n    .split(',')\n    .map(t => t.trim().toLowerCase().replace(/[^a-z0-9-]/g, ''))\n    .filter(t => t.length > 0 && t.length < 30)\n    .slice(0, 5);\n}\n\nreturn [{\n  json: {\n    topic_id: articleData.topic_id,\n    user_id: articleData.user_id,\n    title: articleData.title,\n    url: articleData.url,\n    content: articleData.content,\n    summary: articleData.summary,\n    source: articleData.source,\n    published_date: articleData.published_at,\n    relevance_score: articleData.relevance_score,\n    tags: tags,\n    read_status: 'unread',\n    bookmarked: false\n  }\n}];"
      },
      "id": "extract-tags",
      "name": "Extract Tags",
      "type": "n8n-nodes-base.code",
      "typeVersion": 2,
      "position": [
        3350,
        200
      ]
    },
    {
      "parameters": {
        "method": "POST",
        "url": "http://kairos-rest:3000/articles",
        "sendHeaders": true,
        "headerParameters": {
          "parameters": [
            {
              "name": "apikey",
              "value": "CHANGE_ME_SERVICE_ROLE_KEY"
            },
            {
              "name": "Authorization",
              "value": "Bearer CHANGE_ME_SERVICE_ROLE_KEY"
            },
            {
              "name": "Content-Type",
              "value": "application/json"
            },
            {
              "name": "Prefer",
              "value": "resolution=ignore-duplicates,return=representation"
            }
          ]
        },
        "sendBody": true,
        "specifyBody": "json",
        "jsonBody": "={{ JSON.stringify($json) }}",
        "options": {}
      },
      "id": "insert-article",
      "name": "Insert Article",
      "type": "n8n-nodes-base.httpRequest",
      "typeVersion": 4.2,
      "position": [
        3570,
        200
      ],
      "onError": "continueRegularOutput"
    },
    {
      "parameters": {
        "amount": 2,
        "unit": "seconds"
      },
      "id": "wait",
      "name": "Wait",
      "type": "n8n-nodes-base.wait",
      "typeVersion": 1.1,
      "position": [
        3790,
        200
      ]
    },
    {
      "parameters": {},
      "id": "no-topics",
      "name": "No Topics",
      "type": "n8n-nodes-base.noOp",
      "typeVersion": 1,
      "position": [
        490,
        500
      ]
    },
    {
      "parameters": {},
      "id": "already-exists",
      "name": "Already Exists",
      "type": "n8n-nodes-base.noOp",
      "typeVersion": 1,
      "position": [
        1810,
        400
      ]
    },
    {
      "parameters": {
        "content": "## Kairos RSS Processor v3\n\n### Workflow:\n1. **Trigger**: Schedule (1h) ou Webhook\n2. **Get Topics**: Recupere les topics actifs\n3. **Extract Feeds**: Extrait les URLs RSS\n4. **Fetch RSS**: Lit les flux RSS\n5. **Parse Articles**: Parse et nettoie\n6. **Check Duplicate**: Verifie si l'article existe\n7. **Jina Extract**: Extrait le contenu complet\n8. **IA Analysis** (sequentiel):\n   - Resume (2-3 phrases)\n   - Pertinence (0-100)\n   - Tags (3-5 tags)\n9. **Insert Article**: Sauvegarde en BDD",
        "height": 340,
        "width": 320
      },
      "id": "sticky-note",
      "name": "Sticky Note",
      "type": "n8n-nodes-base.stickyNote",
      "typeVersion": 1,
      "position": [
        -500,
        200
      ]
    }
  ],
  "connections": {
    "Toutes les heures": {
      "main": [
        [
          {
            "node": "Get Topics",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Webhook": {
      "main": [
        [
          {
            "node": "Get Topics",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Get Topics": {
      "main": [
        [
          {
            "node": "Extract Feeds",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Extract Feeds": {
      "main": [
        [
          {
            "node": "Fetch RSS",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Fetch RSS": {
      "main": [
        [
          {
            "node": "Parse Articles",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Parse Articles": {
      "main": [
        [
          {
            "node": "Loop Articles",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Loop Articles": {
      "main": [
        [],
        [
          {
            "node": "Check Duplicate",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Check Duplicate": {
      "main": [
        [
          {
            "node": "Nouvel article?",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Nouvel article?": {
      "main": [
        [
          {
            "node": "Jina Extract",
            "type": "main",
            "index": 0
          }
        ],
        [
          {
            "node": "Already Exists",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Already Exists": {
      "main": [
        [
          {
            "node": "Loop Articles",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Jina Extract": {
      "main": [
        [
          {
            "node": "Merge Content",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Merge Content": {
      "main": [
        [
          {
            "node": "IA - Resume",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "IA - Resume": {
      "main": [
        [
          {
            "node": "Extract Summary",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Extract Summary": {
      "main": [
        [
          {
            "node": "IA - Pertinence",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "IA - Pertinence": {
      "main": [
        [
          {
            "node": "Extract Relevance",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Extract Relevance": {
      "main": [
        [
          {
            "node": "IA - Tags",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "IA - Tags": {
      "main": [
        [
          {
            "node": "Extract Tags",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Extract Tags": {
      "main": [
        [
          {
            "node": "Insert Article",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Insert Article": {
      "main": [
        [
          {
            "node": "Wait",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Wait": {
      "main": [
        [
          {
            "node": "Loop Articles",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  },
  "settings": {
    "executionOrder": "v1"
  },
  "active": true
}

About this workflow

Kairos - RSS Processor v3. Uses httpRequest, rssFeedRead. Scheduled trigger; 23 nodes.

Source: https://github.com/AnythingLegalConsidered/Kairos/blob/55ddb84198e9f34bfdc99435aaf72d322d27331d/n8n/workflows/rss_processor.json — original creator credit. Request a take-down →

More Web Scraping workflows → · Browse all categories →

Related workflows

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