{
  "schema": "kallipolis.resource-quality.v1",
  "id": "resource.safety.owasp-agentic-security-evidence",
  "version": "1.0.0",
  "title": "OWASP agentic AI security evidence checklist",
  "summary": "Publisher-source guidance for turning OWASP agentic risk references into system-specific threat models, least-privilege controls, approval gates, adversarial tests, and monitored evidence rather than a security badge.",
  "content": {
    "kind": "implementation-checklist",
    "facts": [
      {
        "id": "fact.owasp.starting-point",
        "statement": "OWASP describes its Top 10 for Agentic Applications 2026 as a starting point for reducing agentic risk; using the list is not itself a certification or proof that an Agent is secure.",
        "normativeLevel": "informative",
        "sourceIds": ["source.owasp.agentic-top10-2026"]
      },
      {
        "id": "fact.owasp.attack-surfaces",
        "statement": "Agentic threat modelling should cover reasoning, memory, tools, identity, human oversight, and multi-Agent interactions rather than testing only the language model endpoint.",
        "normativeLevel": "informative",
        "sourceIds": ["source.owasp.agentic-navigator"]
      },
      {
        "id": "fact.owasp.prompt-injection-controls",
        "statement": "OWASP states that prompt injection can be direct or indirect and has no known fool-proof prevention; mitigations include output validation, least privilege, human approval for high-risk actions, segregation of external content, and adversarial testing.",
        "normativeLevel": "informative",
        "sourceIds": ["source.owasp.prompt-injection-2025"]
      },
      {
        "id": "fact.owasp.evidence-not-badge",
        "statement": "An Agent security claim should identify the relevant threat, implemented control, test case, result, scope, and observation time; a checklist reference without system-specific evidence does not demonstrate control effectiveness.",
        "normativeLevel": "informative",
        "sourceIds": ["source.owasp.agentic-guide", "source.owasp.agentic-top10-2026"]
      }
    ]
  },
  "language": "en",
  "category": "ai-safety-provenance-and-audit",
  "state": "prepared",
  "canonicalUrl": "https://platoskallipolis.com/resources/ai-safety-provenance-and-audit/owasp-agentic-security-evidence.json",
  "contentHash": "sha256:f59520d39c8a6fcf5251406cf6b5f51890c9b2ab7446b6fa8d62fc3063f1c565",
  "owner": {
    "id": "platos-kallipolis",
    "name": "Plato's Kallipolis",
    "type": "organization",
    "accountabilityUrl": "https://platoskallipolis.com/about"
  },
  "access": {
    "mode": "public",
    "url": "https://platoskallipolis.com/resources/ai-safety-provenance-and-audit/owasp-agentic-security-evidence.json",
    "authentication": "none",
    "robotsAllowed": true
  },
  "provenance": {
    "method": "publisher-declared",
    "sources": [
      {
        "id": "source.owasp.agentic-top10-2026",
        "url": "https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/",
        "title": "OWASP Top 10 for Agentic Applications 2026",
        "publisher": "OWASP GenAI Security Project",
        "observedAt": "2026-08-26T15:38:00Z"
      },
      {
        "id": "source.owasp.agentic-navigator",
        "url": "https://genai.owasp.org/resource/owasp-gen-ai-security-project-agentic-threats-navigator/",
        "title": "OWASP Agentic Threats Navigator",
        "publisher": "OWASP GenAI Security Project",
        "observedAt": "2026-08-26T15:38:00Z"
      },
      {
        "id": "source.owasp.prompt-injection-2025",
        "url": "https://genai.owasp.org/llmrisk/llm01-prompt-injection/",
        "title": "OWASP LLM01 2025 Prompt Injection",
        "publisher": "OWASP GenAI Security Project",
        "observedAt": "2026-08-26T15:38:00Z"
      },
      {
        "id": "source.owasp.agentic-guide",
        "url": "https://genai.owasp.org/resource/securing-agentic-applications-guide-1-0/",
        "title": "OWASP Securing Agentic Applications Guide 1.0",
        "publisher": "OWASP GenAI Security Project",
        "observedAt": "2026-08-26T15:38:00Z"
      }
    ]
  },
  "observedAt": "2026-08-26T15:38:00Z",
  "freshness": {
    "policy": "max-age",
    "state": "current",
    "nextReviewAt": "2026-09-02T15:38:00Z",
    "maxAgeSeconds": 604800
  },
  "license": {
    "id": "cc-by-sa-4.0",
    "url": "https://creativecommons.org/licenses/by-sa/4.0/",
    "redistributionAllowed": true
  },
  "evidence": [
    {
      "id": "evidence.owasp.agentic-security",
      "type": "source-claim",
      "url": "https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/",
      "observedAt": "2026-08-26T15:38:00Z"
    }
  ]
}