{"product_id":"n8n-ai-fact-checker-ollama-wikipedia-score-pass-flag","title":"n8n AI Fact Checker: Ollama + Wikipedia Score PASS\/FLAG","description":"\u003ch3\u003eGet factual PASS\/FLAG results for every AI claim—verified against your facts and Wikipedia\u003c\/h3\u003e\n\u003cp\u003eThis \u003cstrong\u003en8n AI Fact Checker\u003c\/strong\u003e workflow takes any AI output you send to it and returns a strict \u003cstrong\u003ePASS, FLAG, or BLOCK\u003c\/strong\u003e verdict for each factual claim—backed by \u003cstrong\u003equotes from your trusted facts and Wikipedia\u003c\/strong\u003e. It runs on \u003cstrong\u003elocal Ollama\u003c\/strong\u003e (no API key), fails closed when the model is down, and emails a \u003cstrong\u003eweekly truth score\u003c\/strong\u003e.\u003c\/p\u003e\n\n\u003ch3\u003eWhat this workflow does\u003c\/h3\u003e\n\u003cul\u003e\n  \u003cli\u003e\n\u003cstrong\u003eReceives AI output\u003c\/strong\u003e via a \u003cem\u003ePOST webhook\u003c\/em\u003e (or uses a built-in sample) and reads all settings from \u003cstrong\u003eRead Request \u0026amp; Config\u003c\/strong\u003e.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eExtracts checkable claims\u003c\/strong\u003e using a local Ollama model: it targets \u003cstrong\u003enames, numbers, dates, and events\u003c\/strong\u003e while ignoring opinions.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eCollects evidence\u003c\/strong\u003e for each claim by first matching your \u003cstrong\u003etrusted facts\u003c\/strong\u003e from the \u003cstrong\u003eData Table \u003ccode\u003etruth_facts\u003c\/code\u003e\u003c\/strong\u003e, then pulling matching extracts from \u003cstrong\u003eWikipedia\u003c\/strong\u003e. The search subject comes from the claim text (not the model).\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eJudges each claim with citations\u003c\/strong\u003e: a judge model rates each claim \u003cstrong\u003eVERIFIED, UNSUPPORTED, or CONTRADICTED\u003c\/strong\u003e and must include a quote for the verdict.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eEnforces strict validation\u003c\/strong\u003e in code: every quote must exist word-for-word in the evidence (including numbers). If the model fails, results become \u003cstrong\u003eFLAG\u003c\/strong\u003e with a truth score of \u003cstrong\u003e0\u003c\/strong\u003e.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eProduces PASS\/FLAG\/BLOCK\u003c\/strong\u003e: contradictions cause \u003cstrong\u003eBLOCK\u003c\/strong\u003e, too many unsupported claims cause \u003cstrong\u003eFLAG\u003c\/strong\u003e, otherwise \u003cstrong\u003ePASS\u003c\/strong\u003e.\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eLogs and notifies\u003c\/strong\u003e: returns verdicts, records checks in \u003cstrong\u003e\u003ccode\u003etruth_checks\u003c\/code\u003e\u003c\/strong\u003e, and emails alerts on \u003cstrong\u003eFLAG\/BLOCK\u003c\/strong\u003e. Every Monday at \u003cstrong\u003e08:00\u003c\/strong\u003e, it emails a weekly truth score per AI source and deletes checks older than \u003cstrong\u003e90 days\u003c\/strong\u003e.\u003c\/li\u003e\n\u003c\/ul\u003e\n\n\u003ch3\u003eUse cases\u003c\/h3\u003e\n\u003cul\u003e\n  \u003cli\u003eValidate factual sections in AI-generated customer emails or SaaS support answers.\u003c\/li\u003e\n  \u003cli\u003eRisk-control for content workflows: automatically flag unverifiable claims before publishing.\u003c\/li\u003e\n  \u003cli\u003eMonitor reliability over time with a weekly truth score by AI source.\u003c\/li\u003e\n\u003c\/ul\u003e\n\n\u003ch3\u003eTechnical details (n8n)\u003c\/h3\u003e\n\u003cul\u003e\n  \u003cli\u003e\n\u003cstrong\u003eWebhook\u003c\/strong\u003e (POST) for incoming AI output\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eOllama (local)\u003c\/strong\u003e for claim extraction and judging (no external API key)\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eData tables\u003c\/strong\u003e: \u003ccode\u003etruth_facts\u003c\/code\u003e and \u003ccode\u003etruth_checks\u003c\/code\u003e\n\u003c\/li\u003e\n  \u003cli\u003e\n\u003cstrong\u003eNodes used\u003c\/strong\u003e: \u003ccode\u003eif\u003c\/code\u003e, \u003ccode\u003ecode\u003c\/code\u003e, \u003ccode\u003ewebhook\u003c\/code\u003e, \u003ccode\u003edata table\u003c\/code\u003e, \u003ccode\u003eemail send\u003c\/code\u003e, \u003ccode\u003esticky note\u003c\/code\u003e\n\u003c\/li\u003e\n\u003c\/ul\u003e","brand":"N8N Commerce","offers":[{"title":"Default Title","offer_id":46153931751603,"sku":"N8N-20327","price":30.99,"currency_code":"GBP","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0749\/6279\/6723\/files\/YzYJqL6wrjrV1QSoYbQAJ_YuQo5qp5.png?v=1790931702","url":"https:\/\/buyflowscripts.com\/products\/n8n-ai-fact-checker-ollama-wikipedia-score-pass-flag","provider":"N8N Commerce","version":"1.0","type":"link"}