{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://spec.falsify.dev/schema/prml-v0.2.schema.json",
  "title": "PRML v0.2 — Pre-Registered ML Manifest (RFC, freeze 2026-05-22)",
  "description": "Formal JSON Schema for the PRML v0.2 RFC (https://spec.falsify.dev/v0.2-rfc). v0.2 adds optional fields (streaming variant, runner attestation, revocation) over v0.1. Every v0.1 manifest is a valid v0.2 manifest. Fully backwards-compatible at the hash level. CC BY 4.0.",
  "type": "object",
  "additionalProperties": false,
  "required": [
    "version",
    "claim_id",
    "metric",
    "comparator",
    "threshold",
    "dataset",
    "seed",
    "producer"
  ],
  "properties": {
    "version": {
      "type": "string",
      "enum": ["prml/0.1", "prml/0.2"],
      "description": "Specification version. v0.2 verifiers MUST accept v0.1 manifests; v0.1 verifiers MAY reject v0.2-only fields."
    },
    "claim_id": {
      "type": "string",
      "pattern": "^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-7[0-9a-fA-F]{3}-[89ab][0-9a-fA-F]{3}-[0-9a-fA-F]{12}$"
    },
    "created_at": {
      "type": "string",
      "format": "date-time",
      "description": "RFC 3339 timestamp. Required for static-mode manifests; replaced by pre_registered_from/to in streaming mode."
    },
    "metric": {"type": "string", "minLength": 1, "maxLength": 256},
    "comparator": {"type": "string", "enum": [">=", "<=", ">", "<", "=="]},
    "threshold": {"type": "number"},
    "dataset": {
      "type": "object",
      "additionalProperties": false,
      "required": ["id", "hash"],
      "properties": {
        "id": {"type": "string", "minLength": 1},
        "hash": {"type": "string"},
        "uri": {"type": "string", "format": "uri"}
      }
    },
    "seed": {
      "oneOf": [
        {"type": "integer", "minimum": 0, "maximum": 18446744073709551615},
        {"type": "integer", "minimum": -9223372036854775808, "maximum": -1},
        {"type": "null"}
      ]
    },
    "producer": {
      "type": "object",
      "additionalProperties": false,
      "required": ["id"],
      "properties": {
        "id": {"type": "string", "minLength": 1},
        "signature": {"type": "string"}
      }
    },

    "model": {
      "type": "object",
      "additionalProperties": false,
      "properties": {
        "id": {"type": "string", "minLength": 1},
        "hash": {"type": "string"},
        "uri": {"type": "string", "format": "uri"}
      }
    },

    "compute_envelope": {
      "type": "object",
      "additionalProperties": true
    },
    "prior_hash": {
      "type": "string",
      "pattern": "^[0-9a-fA-F]{64}$"
    },
    "notes": {"type": "string", "maxLength": 4096},

    "prml_mode": {
      "type": "string",
      "enum": ["static", "streaming"],
      "description": "v0.2 P-01. Default 'static' (v0.1 behaviour). 'streaming' enables window-based commitments."
    },

    "value_method": {
      "type": "string",
      "minLength": 1,
      "maxLength": 256,
      "description": "v0.2 P-01. Aggregation rule identifier in streaming mode (e.g. 'lmsys_anonymous_chat_arena_v1', 'mean_over_window')."
    },

    "sample_size": {
      "type": "integer",
      "minimum": 0,
      "description": "v0.2 P-01. In static mode: exact eval set size. In streaming mode: minimum sample size below which the verdict is undefined."
    },

    "pre_registered_from": {
      "type": "string",
      "format": "date-time",
      "description": "v0.2 P-01. Streaming mode start of window."
    },

    "pre_registered_to": {
      "type": "string",
      "format": "date-time",
      "description": "v0.2 P-01. Streaming mode end of window."
    },

    "runner_attestation": {
      "type": "string",
      "minLength": 1,
      "description": "v0.2 P-02. Opaque URI to an out-of-band execution attestation (Sigstore, in-toto, TEE). PRML records that an attestation was emitted; the URI is not interpreted by canonicalisation."
    },

    "revoked_at": {
      "type": "string",
      "format": "date-time",
      "description": "v0.2 P-03. Timestamp at which the publisher revoked this manifest. The hash still verifies; verifiers MUST surface revocation status separately from hash status."
    },

    "revocation_reason": {
      "type": "string",
      "enum": ["dataset_compromised", "model_recalled", "author_request", "other"],
      "description": "v0.2 P-03. Controlled vocabulary for the reason the publisher revoked the manifest."
    }
  },

  "allOf": [
    {
      "if": {"properties": {"prml_mode": {"const": "streaming"}}, "required": ["prml_mode"]},
      "then": {
        "required": ["pre_registered_from", "pre_registered_to", "value_method", "sample_size"]
      }
    },
    {
      "if": {"required": ["revoked_at"]},
      "then": {"required": ["revocation_reason"]}
    },
    {
      "if": {"properties": {"prml_mode": {"const": "static"}}, "required": ["prml_mode"]},
      "then": {"required": ["created_at"]}
    },
    {
      "if": {"not": {"required": ["prml_mode"]}},
      "then": {"required": ["created_at"]}
    }
  ],

  "examples": [
    {
      "version": "prml/0.2",
      "claim_id": "01900000-0000-7000-8000-000000000014",
      "prml_mode": "streaming",
      "metric": "elo_rating",
      "value_method": "lmsys_anonymous_chat_arena_v1",
      "comparator": ">=",
      "threshold": 1300,
      "dataset": {"id": "lmsys-arena-live", "hash": "n/a-streaming"},
      "seed": null,
      "producer": {"id": "studio-11.co"},
      "model": {"id": "claude-3.5-sonnet@2025-10-01"},
      "sample_size": 1000,
      "pre_registered_from": "2026-05-01T00:00:00Z",
      "pre_registered_to": "2026-06-01T00:00:00Z"
    },
    {
      "version": "prml/0.2",
      "claim_id": "01900000-0000-7000-8000-000000000016",
      "created_at": "2026-04-01T12:00:00Z",
      "metric": "accuracy",
      "comparator": ">=",
      "threshold": 0.92,
      "dataset": {"id": "imagenet-val-2012", "hash": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"},
      "seed": 42,
      "producer": {"id": "studio-11.co"},
      "revoked_at": "2026-05-15T10:00:00Z",
      "revocation_reason": "dataset_compromised"
    }
  ]
}