{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "$id": "https://raw.githubusercontent.com/api-evangelist/sensible-so/main/json-schema/sensible-so-extraction-schema.json",
  "title": "SensibleExtraction",
  "description": "An asynchronous-extraction record returned by the Sensible API. Captures status, parsed_document, validations, and source linkage.",
  "type": "object",
  "required": ["id", "type", "status"],
  "properties": {
    "id": {
      "type": "string",
      "format": "uuid",
      "description": "Unique extraction identifier returned by Sensible."
    },
    "type": {
      "type": "string",
      "description": "Document type slug (e.g. bank_statement, acord_25, rate_confirmation)."
    },
    "configuration": {
      "type": "string",
      "description": "Name of the SenseML configuration that produced the extraction."
    },
    "environment": {
      "type": "string",
      "enum": ["production", "development"],
      "description": "Sensible publication environment that ran the config."
    },
    "status": {
      "type": "string",
      "enum": ["WAITING", "PROCESSING", "FAILED", "COMPLETE"],
      "description": "Current extraction status. Poll until COMPLETE or FAILED."
    },
    "created_at": { "type": "string", "format": "date-time" },
    "updated_at": { "type": "string", "format": "date-time" },
    "document_name": { "type": ["string", "null"] },
    "download_link": {
      "type": ["string", "null"],
      "format": "uri",
      "description": "Pre-signed S3 URL for downloading the source document."
    },
    "parsed_document": {
      "type": ["object", "null"],
      "additionalProperties": {
        "$ref": "#/definitions/ParsedField"
      },
      "description": "Field-name → parsed field. Null until extraction completes."
    },
    "validations": {
      "type": ["array", "null"],
      "items": { "$ref": "#/definitions/Validation" }
    },
    "review_status": {
      "type": ["string", "null"],
      "enum": [null, "UNREVIEWED", "IN_REVIEW", "APPROVED", "REJECTED"]
    },
    "coverage": {
      "type": ["number", "null"],
      "minimum": 0,
      "maximum": 1,
      "description": "Fraction of fields in the config that returned a non-null value."
    }
  },
  "definitions": {
    "ParsedField": {
      "type": "object",
      "required": ["type"],
      "properties": {
        "type": {
          "type": "string",
          "enum": ["string", "number", "integer", "boolean", "date", "currency", "address", "name", "phone_number", "row", "table"]
        },
        "value": {
          "description": "Parsed value (typed) or null if not found."
        },
        "source": {
          "type": "object",
          "description": "Optional source coordinates / line traceability."
        },
        "confidence": {
          "type": "object",
          "description": "Optional confidence scoring for LLM-based methods."
        }
      }
    },
    "Validation": {
      "type": "object",
      "required": ["severity"],
      "properties": {
        "description": { "type": "string" },
        "severity": { "type": "string", "enum": ["error", "warning", "info"] },
        "fields": { "type": "array", "items": { "type": "string" } },
        "passed": { "type": "boolean" }
      }
    }
  }
}