{
 "id": "AIR-2026-0010",
 "slug": "langflow-smart-transform-prompt-injection-and-code-execution",
 "title": "Langflow Smart Transform prompt injection and code execution vulnerability",
 "summary": "Langflow contained a vulnerability in the LambdaFilterComponent evaluation function that allowed remote code injection via prompt manipulation. This advisory was marked as a duplicate of another tracking record.",
 "type": "attack_on_ai",
 "categories": [
  "prompt_injection",
  "infra_vuln"
 ],
 "severity": 2,
 "targets": [
  "framework_or_tooling"
 ],
 "products": [
  "Langflow"
 ],
 "incident_date": "2026-05-03",
 "incident_date_precision": "day",
 "first_reported": "2026-05-03",
 "sources": [
  {
   "url": "https://github.com/advisories/GHSA-hhxr-7j22-694r",
   "publisher": "GitHub Advisory Database",
   "title": "GHSA-hhxr-7j22-694r: Duplicate Advisory: Prompt injection in Langflow Smart Transform can lead to code execution",
   "published_at": "2026-05-03"
  }
 ],
 "source_count": 1,
 "confidence": 0.95,
 "status": "published",
 "origin": "seed_ghsa",
 "tags": [
  "langflow",
  "prompt-injection",
  "code-injection"
 ],
 "created_at": "2026-10-18T06:33:31.967Z",
 "updated_at": "2026-10-18T06:33:31.967Z"
}