{
  "$schema": "http://json-schema.org/draft-07/schema#",
  "$id": "https://stellaops.io/schemas/corpus-sample.v1.json",
  "title": "CorpusSample",
  "description": "Schema for ground-truth corpus samples used in reachability benchmarking",
  "type": "object",
  "required": ["sampleId", "name", "format", "arch", "sinks"],
  "properties": {
    "sampleId": {
      "type": "string",
      "pattern": "^gt-[0-9]{4}$",
      "description": "Unique identifier for the sample (e.g., gt-0001)"
    },
    "name": {
      "type": "string",
      "description": "Human-readable name for the sample"
    },
    "description": {
      "type": "string",
      "description": "Detailed description of what this sample tests"
    },
    "category": {
      "type": "string",
      "enum": ["basic", "indirect", "stripped", "obfuscated", "guarded", "callback", "virtual"],
      "description": "Sample category for organization"
    },
    "format": {
      "type": "string",
      "enum": ["elf64", "elf32", "pe64", "pe32", "macho64", "macho32"],
      "description": "Binary format"
    },
    "arch": {
      "type": "string",
      "enum": ["x86_64", "x86", "aarch64", "arm32", "riscv64"],
      "description": "Target architecture"
    },
    "language": {
      "type": "string",
      "enum": ["c", "cpp", "rust", "go"],
      "description": "Source language (for reference)"
    },
    "compiler": {
      "type": "object",
      "properties": {
        "name": { "type": "string" },
        "version": { "type": "string" },
        "flags": { "type": "array", "items": { "type": "string" } }
      },
      "description": "Compiler information used to build the sample"
    },
    "entryPoint": {
      "type": "string",
      "default": "main",
      "description": "Entry point function name"
    },
    "sinks": {
      "type": "array",
      "minItems": 1,
      "items": {
        "type": "object",
        "required": ["sinkId", "signature", "expected"],
        "properties": {
          "sinkId": {
            "type": "string",
            "pattern": "^sink-[0-9]{3}$",
            "description": "Unique sink identifier within the sample"
          },
          "signature": {
            "type": "string",
            "description": "Function signature of the sink"
          },
          "sinkType": {
            "type": "string",
            "enum": ["memory_corruption", "command_injection", "sql_injection", "path_traversal", "format_string", "crypto_weakness", "custom"],
            "description": "Type of vulnerability represented by the sink"
          },
          "expected": {
            "type": "string",
            "enum": ["reachable", "unreachable", "conditional"],
            "description": "Expected reachability determination"
          },
          "expectedPaths": {
            "type": "array",
            "items": {
              "type": "array",
              "items": { "type": "string" }
            },
            "description": "Expected call paths from entry to sink (for reachable sinks)"
          },
          "guardConditions": {
            "type": "array",
            "items": {
              "type": "object",
              "properties": {
                "variable": { "type": "string" },
                "condition": { "type": "string" },
                "value": { "type": "string" }
              }
            },
            "description": "Guard conditions that protect the sink (for conditional sinks)"
          },
          "notes": {
            "type": "string",
            "description": "Additional notes about this sink"
          }
        }
      },
      "description": "List of sinks with expected reachability"
    },
    "metadata": {
      "type": "object",
      "properties": {
        "createdAt": { "type": "string", "format": "date-time" },
        "createdBy": { "type": "string" },
        "version": { "type": "string" },
        "sha256": { "type": "string", "pattern": "^[a-f0-9]{64}$" }
      },
      "description": "Metadata about the sample"
    }
  }
}