{
  "schemaVersion": 1,
  "generatedBy": "manual-benchmark-evolution-v1",
  "observedAt": "2026-05-30",
  "canonicalRule": "A benchmark claim is active only when corpus digest, scoring rubric, judge panel, receipts, and CI validation are present. competition-math benchmark claims must be raw-score claims.",
  "publication": {
    "canonicalSource": "https://github.com/szl-holdings/a11oy",
    "huggingFaceResultsRepo": "SZLHOLDINGS/a11oy-test-results",
    "huggingFaceRepoType": "dataset",
    "publishMode": "mirror-not-canonical"
  },
  "entries": [
    {
      "id": "mathcomp-raw-score-v1",
      "family": "mathematics",
      "status": "staged-no-corpus-digest",
      "sourceEcosystem": [
        "agi-forecast",
        "lutar-lean",
        "a11oy"
      ],
      "corpus": {
        "corpusId": "mathcomp-public-years",
        "corpusVersion": "v1",
        "immutability": "content-addressed",
        "sourceUri": "external-official-or-licensed-source",
        "license": "do-not-redistribute-unless-permitted",
        "canonicalization": "utf8-nfc-jsonl-sorted-keys",
        "digestStatus": "external-digest-required-before-scoring",
        "problemCount": 0,
        "sealed": false
      },
      "scoring": {
        "scoreType": "raw_points",
        "maxRawScore": 120,
        "perProblemMax": 10,
        "partialCreditAllowed": true,
        "reportFields": [
          "year",
          "problem",
          "earned",
          "possible",
          "judgeIds",
          "disagreement"
        ]
      },
      "honesty": {
        "allowedClaim": "staged competition-math raw-score harness design; no live score claimed",
        "disallowedClaims": [
          "solved the benchmark",
          "beat the benchmark",
          "AGI proven"
        ],
        "headlineThresholdRequiresUnanimousJudges": true
      },
      "formulaRoutes": [
        {
          "formula": "FalsePosition",
          "theoremRuntimeManifestId": "RUNTIME-FALSE-POSITION",
          "allowedUse": "routing-or-verification-when-problem-matches"
        },
        {
          "formula": "MadhavaBound",
          "theoremRuntimeManifestId": "RUNTIME-MADHAVA-BOUND",
          "allowedUse": "numeric-bound-check-only"
        },
        {
          "formula": "LiuHuiPi",
          "theoremRuntimeManifestId": "RUNTIME-LIU-HUI-PI",
          "allowedUse": "numeric-approximation-check-only"
        },
        {
          "formula": "SummationInvariant",
          "theoremRuntimeManifestId": "RUNTIME-SUMMATION-INVARIANT",
          "allowedUse": "additive-invariant-check-only"
        },
        {
          "formula": "QECLineage",
          "theoremRuntimeManifestId": "RUNTIME-QEC-LINEAGE",
          "allowedUse": "receipt-integrity-not-math-credit"
        }
      ],
      "judges": [
        {
          "id": "raw_grader",
          "role": "rubric_score",
          "required": true
        },
        {
          "id": "proof_judge",
          "role": "proof_validity",
          "required": true
        },
        {
          "id": "provenance_judge",
          "role": "receipt_and_claim_guard",
          "required": true
        }
      ],
      "receipts": {
        "required": true,
        "ledgerPathPattern": "benchmarks/results/{runId}/receipts.jsonl",
        "chain": "hash_chain",
        "quorum": "2-of-3"
      },
      "ciGates": [
        "validate-benchmark-map",
        "verify-corpus-digests",
        "verify-receipt-chain",
        "verify-formula-routes",
        "reject-unsupported-benchmark-claims"
      ]
    }
  ]
}