{
  "id": "guardian-falsification",
  "type": "entity",
  "name": "Guardian falsification",
  "summary": "A discipline for proving that an automated check (a \"guardian\") can actually fail: deliberately plant a defect, watch the guardian go red, restore the correct state, watch it go green. A guardian that has never been observed red is an assumption, not a control. Practiced in production in the Citarium framework, where every validator check family traces back to a real incident.",
  "locale": "en",
  "tags": [
    "testing",
    "guardians",
    "falsification",
    "ci",
    "regression"
  ],
  "relations": [
    {
      "rel": "alarmed_by",
      "target": "check-counter-alarm"
    },
    {
      "rel": "explained_in",
      "target": "falsify-your-first-guardian"
    },
    {
      "rel": "related",
      "target": "behavioral-canaries"
    }
  ],
  "questions": [
    "How do I know my CI checks would actually catch a failure?",
    "What is guardian falsification?",
    "Why should I make my own checks fail on purpose?"
  ],
  "claims": [
    {
      "id": "c1",
      "text": "Guardian falsification validates a check by planting a known defect, confirming the guardian turns red, restoring the correct state, and confirming it turns green again; only a guardian that has been observed failing counts as evidence rather than decoration.",
      "sources": [
        "citarium-methodology-brief"
      ],
      "confidence": 0.95
    },
    {
      "id": "c2",
      "text": "The Citarium content validator is deployed as each instance's first guardian and runs from CI on day one; its source states that every check exists because an incident paid for it.",
      "sources": [
        "citarium-core-validator"
      ],
      "confidence": 0.95
    },
    {
      "id": "c3",
      "text": "The falsified-guardian methodology is a living case study with real incident history in the Citarium framework, not a thought experiment.",
      "sources": [
        "citarium-methodology-brief"
      ],
      "confidence": 0.9
    }
  ],
  "takeaways": [],
  "faqs": [],
  "evidence_tier": "primary",
  "evidence": {
    "level": "production",
    "source_types": [
      "production_system",
      "personal_experience"
    ]
  },
  "moat_flag": true,
  "winning_edge": "First-hand and reproducible: documented from the framework that runs the methodology in production, with the validator source code as citable evidence. Existing literature covers mutation testing of unit tests, not falsification of content and agent guardians as an operating discipline.",
  "confidence": 0.9,
  "last_verified": "2026-08-08",
  "canonical_url": "https://agentreliability.dev/k/guardian-falsification",
  "api_url": "https://agentreliability.dev/api/k/guardian-falsification.json",
  "jsonld": {
    "@context": "https://schema.org",
    "name": "Guardian falsification",
    "description": "A discipline for proving that an automated check (a \"guardian\") can actually fail: deliberately plant a defect, watch the guardian go red, restore the correct state, watch it go green. A guardian that has never been observed red is an assumption, not a control. Practiced in production in the Citarium framework, where every validator check family traces back to a real incident.",
    "url": "https://agentreliability.dev/k/guardian-falsification",
    "license": "https://spdx.org/licenses/CC-BY-4.0.html",
    "dateModified": "2026-08-08",
    "citation": [
      {
        "@type": "CreativeWork",
        "name": "Citarium content validator source (@citarium/core v0.1.0, compiler/validate.ts)"
      },
      {
        "@type": "CreativeWork",
        "name": "agent-reliability editorial brief and blueprint (Gate 1 approved, 2026-08-08)"
      }
    ],
    "author": {
      "@type": "Person",
      "name": "Santiago Santa María Morales",
      "jobTitle": "practitioner — harness engineering and agent evaluation in production"
    },
    "@type": "Article",
    "headline": "Guardian falsification"
  }
}