{
  "mode": "deterministic_reference",
  "model_calls": 0,
  "human_participants": 0,
  "real_business_execution": false,
  "case_count": 3,
  "numeric_checks": 9,
  "numeric_checks_passed": 9,
  "current_source_sha256": {
    "__init__.py": "5e567e37c2edcc5bbdb510bd5785b55c6ac0bcdf0411a6b0ba9bd790d7f7e627",
    "__main__.py": "3129adff95e0f38b87ab34cddfc27082c7892bcaec267ff5d40f160fdf55b44a",
    "cli.py": "3792b4d311a7dc3a9665d1564b3de8863fdd8d74d2c3d95c0a0b7b4f23fe9bc5",
    "codex_adapter.py": "b2b0c18b930e2c88ea2668554aa428a01e0b5222e7a7be1545185801d5df583f",
    "dashboard.py": "e212178009b3dde9b67471db3d386bd776a6ab017b3cb2d01860fbba8ca4aea0",
    "data.py": "a964c97df5bbd4f265e564e340cf11c6d79abae02687f49eca9f1ce146d3cf21",
    "evaluation.py": "6a58177cb88b8adbbf51a22a4b9d07455589e703b3ac7b5e82b1ec606f36f1d7",
    "fresh.py": "13779bdb12f52de1b85c0056a4db79b02f0f798aad734f39f858b068c330b42e",
    "model.py": "8b1b1c6bd33940f64652e053ef15bfba6294bdee1ed4c0fda5c8f06685c7d448",
    "orchestrator.py": "840fe5c9548e985fc843cb412e06b8f7eef4697f780c97437cfd5f208488b0e5",
    "policy.py": "72f823baf1c5eb41b93b38dd7e4f92a54175614f268b3c346fbc500efb5b8dc8",
    "protocol.py": "dfeef9495fc5d52e7b3d4e642111b081ba73dc17ac390686f28ebe6d605cf27a",
    "provenance.py": "bd2161bb05288defb57efb1475a76dd9f1f711c97c5ec8c7183b8f9464b9ce18",
    "sdk_agent.py": "fce5b1effec8feb0975dc58d63b5db200c3718cfe5dad3e1b3aee1bfe91708e9",
    "tools.py": "51c560ea5f219d93e9ea09158fb477abeb84a3113bcc93f06f52211dbffe1393"
  },
  "design_boundary": "Root-directed synthetic stress cases; separate author did not read implementation. Cases informed the patch and are now regression examples, not independent holdout or human-effectiveness evidence.",
  "results": [
    {
      "case_id": "X01",
      "before_decision": "INVESTIGATE",
      "before_driver": "变化集中在少数门店，应先与同簇门店比较再决定区域策略。",
      "after_decision": "INVESTIGATE",
      "after_driver": "四周试点表现不能覆盖最近一周的贡献转负；续行前需要专项复核。",
      "after_action": {
        "type": "recent_pilot_diagnostic",
        "description": "暂缓提出试点续行或扩大建议；财务核对末周目标商品的单位成本、促销费用凭证与生效日期，区分记录问题和真实经营变化，将结果交品类负责人复核。",
        "scope": {
          "stores": [
            "C1",
            "U1"
          ],
          "skus": [
            "P1"
          ],
          "duration_days": 2
        },
        "requires_human_approval": true,
        "execution_status": "not_executed"
      },
      "owner": [
        "Category Manager",
        "Finance"
      ],
      "review": {
        "due": "2026-09-21",
        "relative_to": "synthetic data as_of",
        "within_days": 2,
        "metrics": [
          "contribution",
          "revenue_vs_budget",
          "stockout_rate",
          "unknown_inventory_loss"
        ]
      },
      "raw_row_check": {
        "four_week_incremental_contribution": "1368.00",
        "latest_week_contribution": "-708.00",
        "latest_week_worst_pool_stockout": "0",
        "latest_week": "2026-09-07"
      },
      "numeric_checks_passed": 3,
      "same_input_sha256": true
    },
    {
      "case_id": "X02",
      "before_decision": "INVESTIGATE",
      "before_driver": "变化集中在少数门店，应先与同簇门店比较再决定区域策略。",
      "after_decision": "INVESTIGATE",
      "after_driver": "四周试点表现不能覆盖最近一周的局部缺货；续行前需要专项复核。",
      "after_action": {
        "type": "recent_pilot_diagnostic",
        "description": "暂缓提出试点续行或扩大建议；门店与供应负责人核对指定渠道的缺货事件、补货时点和当前可售库存，确认能否支持下一周试点，将结果交品类负责人复核。",
        "scope": {
          "stores": [
            "U1"
          ],
          "skus": [
            "P1"
          ],
          "duration_days": 2
        },
        "requires_human_approval": true,
        "execution_status": "not_executed"
      },
      "owner": [
        "Category Manager",
        "Store Ops",
        "Supply"
      ],
      "review": {
        "due": "2026-09-21",
        "relative_to": "synthetic data as_of",
        "within_days": 2,
        "metrics": [
          "contribution",
          "revenue_vs_budget",
          "stockout_rate",
          "unknown_inventory_loss"
        ]
      },
      "raw_row_check": {
        "four_week_incremental_contribution": "882.00",
        "latest_week_contribution": "990.00",
        "latest_week_worst_pool_stockout": "0.2",
        "latest_week": "2026-09-07"
      },
      "numeric_checks_passed": 3,
      "same_input_sha256": true
    },
    {
      "case_id": "X03",
      "before_decision": "INVESTIGATE",
      "before_driver": "变化集中在少数门店，应先与同簇门店比较再决定区域策略。",
      "after_decision": "ACT",
      "after_driver": "合成对照试验显示正增量贡献，当前护栏与明确试点授权同时满足。",
      "after_action": {
        "type": "bounded_pilot",
        "description": "建议在既有合成授权内延续 7 天小范围试点；仅形成建议，不执行价格、订货或促销。",
        "scope": {
          "stores": [
            "C1",
            "U1"
          ],
          "skus": [
            "P1"
          ],
          "duration_days": 7,
          "max_stores": 2,
          "max_spend_cny": 500,
          "authority_basis": "synthetic scenario approval; not real authorization"
        },
        "requires_human_approval": true,
        "execution_status": "not_executed"
      },
      "owner": [
        "Category Manager",
        "Promotion Owner",
        "Store Ops"
      ],
      "review": {
        "due": "2026-09-26",
        "relative_to": "synthetic data as_of",
        "within_days": 7,
        "metrics": [
          "contribution",
          "revenue_vs_budget",
          "stockout_rate",
          "unknown_inventory_loss"
        ]
      },
      "raw_row_check": {
        "four_week_incremental_contribution": "960.00",
        "latest_week_contribution": "1068.00",
        "latest_week_worst_pool_stockout": "0",
        "latest_week": "2026-09-07"
      },
      "numeric_checks_passed": 3,
      "same_input_sha256": true
    }
  ]
}