NODEDC_1C/llm_normalizer/data/eval_cases/eval-WWb7wzqwE2.report.json

1034 lines
37 KiB
JSON
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

{
"run_id": "eval-WWb7wzqwE2",
"timestamp": "2026-03-23T15:31:25.602Z",
"mode": "single-pass-strict",
"use_mock": false,
"prompt_version": "normalizer_v1_1_1",
"dataset": {
"source": "file",
"file": "normalizer_eval_v1_1_30cases.json"
},
"cases_total": 30,
"metrics": {
"schema_validation_pass_rate": 100,
"intent_class_accuracy": 90,
"route_hint_accuracy": 100,
"causal_flag_accuracy": 100,
"high_confidence_error_rate": 3.33
},
"baseline_metrics": {
"schema_validation_pass_rate": 100,
"intent_class_accuracy": 72.73,
"route_hint_accuracy": 90.91,
"causal_flag_accuracy": 81.82,
"high_confidence_error_rate": 9.09
},
"baseline_delta": {
"schema_validation_pass_rate": 0,
"intent_class_accuracy": 17.27,
"route_hint_accuracy": 9.09,
"causal_flag_accuracy": 18.18,
"high_confidence_error_rate": -5.76
},
"class_accuracy": {
"cross_entity": {
"total": 10,
"passed": 10,
"accuracy_percent": 100
},
"heavy_analytical": {
"total": 5,
"passed": 2,
"accuracy_percent": 40
},
"drilldown_explain": {
"total": 5,
"passed": 5,
"accuracy_percent": 100
},
"rule_based_account_control": {
"total": 5,
"passed": 5,
"accuracy_percent": 100
},
"anomaly_probe": {
"total": 2,
"passed": 2,
"accuracy_percent": 100
},
"ambiguous_human_query": {
"total": 1,
"passed": 1,
"accuracy_percent": 100
},
"period_close_risk": {
"total": 2,
"passed": 2,
"accuracy_percent": 100
}
},
"budget": {
"requests_total": 30,
"retries_used": 0,
"guidance": {
"forensic_calls_max": 10,
"final_eval_calls_max": 30,
"target_total_calls_max": 40,
"hard_cap_calls_max": 45
}
},
"mismatches": [
{
"case_id": "NQ-002",
"expected_intent_class": "heavy_analytical",
"actual_intent_class": "period_close_risk",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"comment": "Route chosen correctly, but intent_class drifted into a neighboring taxonomy bucket.",
"trace_id": "TEr2iGCFpFVqWf"
},
{
"case_id": "NQ-007",
"expected_intent_class": "heavy_analytical",
"actual_intent_class": "period_close_risk",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"comment": "Route chosen correctly, but intent_class drifted into a neighboring taxonomy bucket.",
"trace_id": "O3hne6oFNSdmas"
},
{
"case_id": "V11-HA-004",
"expected_intent_class": "heavy_analytical",
"actual_intent_class": "period_close_risk",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": true,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"comment": "Route chosen correctly, but intent_class drifted into a neighboring taxonomy bucket.",
"trace_id": "w-XgZuNlZ_Tl2m"
}
],
"bad_confidence_cases": [
{
"case_id": "NQ-002",
"confidence_overall": "high",
"intent_match": false,
"route_match": true,
"causal_match": true,
"trace_id": "TEr2iGCFpFVqWf"
}
],
"results": [
{
"case_id": "NQ-001",
"raw_question": "По каким поставщикам на конец июня не бьются взаиморасчеты, покажи документы, оплаты и хвосты.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "hSoFChVp3Nx0tp",
"request_count_for_case": 1
},
{
"case_id": "NQ-006",
"raw_question": "По каким реализациям 90/62 хвосты не закрылись оплатой, разложи по цепочке документов.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "_OVB8PGHJHZ_By",
"request_count_for_case": 1
},
{
"case_id": "V11-CE-003",
"raw_question": "Где в июне не сходится 60/51: разложи по документу, оплате и закрывающему, чем подтверждается каждый шаг.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": true
},
"confidence_overall": "medium",
"trace_id": "GoatFCi0d3wuGg",
"request_count_for_case": 1
},
{
"case_id": "V11-CE-004",
"raw_question": "Разложи по контрагентам цепочку: отгрузка -> оплата -> закрывающий, чтобы понять где рвется подтверждение.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": true
},
"confidence_overall": "medium",
"trace_id": "aFq6uvVv0YQEMh",
"request_count_for_case": 1
},
{
"case_id": "V11-CE-005",
"raw_question": "По поставщикам где повисло в цепочке поступление-оплата-закрытие по 60, покажи проблемные связки.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": true,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "dlXfVz6q-NPvZp",
"request_count_for_case": 1
},
{
"case_id": "V11-CE-006",
"raw_question": "Найди где по 62 не собралось: нужен разбор по документам, оплатам и проводкам с причинно-следственной цепочкой.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "ATg-we7BjmmtOv",
"request_count_for_case": 1
},
{
"case_id": "V11-CE-007",
"raw_question": "Покажи по июню все случаи когда реализация без оплаты и где в цепочке ошибка подтверждения.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "9l4qpeD-WwbZV-",
"request_count_for_case": 1
},
{
"case_id": "V11-CE-008",
"raw_question": "Сделай причинный разбор хвостов по 60: документ, оплата, проводка, закрывающий, где пошло криво.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "NTbMSWNRbQngAo",
"request_count_for_case": 1
},
{
"case_id": "V11-CE-009",
"raw_question": "Почему у части покупателей не видно закрытия, разложи цепочку документов и оплат по июню.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "qCOJXwsDRiH1IR",
"request_count_for_case": 1
},
{
"case_id": "V11-CE-010",
"raw_question": "У кого из контрагентов в июне хвосты между 60 и банком, разложи по документам/оплатам/закрывающим.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "cross_entity",
"actual_intent_class": "cross_entity",
"expected_route_hint": "hybrid_store_plus_live",
"actual_route_hint": "hybrid_store_plus_live",
"expected_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true
},
"actual_requires": {
"needs_cross_entity_join": true,
"needs_causal_chain": true,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "wT0KLnMK9bOHSm",
"request_count_for_case": 1
},
{
"case_id": "NQ-002",
"raw_question": "Сделай рейтинг самых рисковых хвостов перед закрытием периода за июнь.",
"validation_passed": true,
"intent_match": false,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "heavy_analytical",
"actual_intent_class": "period_close_risk",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"confidence_overall": "high",
"trace_id": "TEr2iGCFpFVqWf",
"request_count_for_case": 1
},
{
"case_id": "NQ-007",
"raw_question": "Что у нас выглядит самым проблемным перед закрытием июня, если смотреть на компанию в целом?",
"validation_passed": true,
"intent_match": false,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "heavy_analytical",
"actual_intent_class": "period_close_risk",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "O3hne6oFNSdmas",
"request_count_for_case": 1
},
{
"case_id": "V11-HA-003",
"raw_question": "Собери топ-10 риск-зон учета по июню и приоритизируй, куда лезть сначала.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "heavy_analytical",
"actual_intent_class": "heavy_analytical",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "pDkgBWjsWjwl5q",
"request_count_for_case": 1
},
{
"case_id": "V11-HA-004",
"raw_question": "Дай обзорный риск-срез перед сдачей отчетности: где максимальная концентрация ошибок.",
"validation_passed": true,
"intent_match": false,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "heavy_analytical",
"actual_intent_class": "period_close_risk",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": true,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "w-XgZuNlZ_Tl2m",
"request_count_for_case": 1
},
{
"case_id": "V11-HA-005",
"raw_question": "Сделай приоритизированный обзор ручных проверок по компании за июнь.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "heavy_analytical",
"actual_intent_class": "heavy_analytical",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "VLc-GocRaKBZ_S",
"request_count_for_case": 1
},
{
"case_id": "NQ-003",
"raw_question": "Покажи документ по номеру 000123 и строку проводки, нужен точный source-of-record.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "drilldown_explain",
"actual_intent_class": "drilldown_explain",
"expected_route_hint": "live_mcp_drilldown",
"actual_route_hint": "live_mcp_drilldown",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": true,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": true,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "jk9jJrWtc58iwt",
"request_count_for_case": 1
},
{
"case_id": "NQ-008",
"raw_question": "Покажи по банку документ №TRX-88 и связанную проводку по 51.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "drilldown_explain",
"actual_intent_class": "drilldown_explain",
"expected_route_hint": "live_mcp_drilldown",
"actual_route_hint": "live_mcp_drilldown",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": true,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": true,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "yASMRehqOtv9pO",
"request_count_for_case": 1
},
{
"case_id": "V11-DD-003",
"raw_question": "Покажи проводку по документу INV-2020-0615, нужна конкретная строка и источник.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "drilldown_explain",
"actual_intent_class": "drilldown_explain",
"expected_route_hint": "live_mcp_drilldown",
"actual_route_hint": "live_mcp_drilldown",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": true,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": true,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "sdYzBBUwY3gVLq",
"request_count_for_case": 1
},
{
"case_id": "V11-DD-004",
"raw_question": "Дай точечный drilldown по документу №PAY-441 и его проводке по 51.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "drilldown_explain",
"actual_intent_class": "drilldown_explain",
"expected_route_hint": "live_mcp_drilldown",
"actual_route_hint": "live_mcp_drilldown",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": true,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": true,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "mdMfZ_rp7AYQxN",
"request_count_for_case": 1
},
{
"case_id": "V11-DD-005",
"raw_question": "Покажи карточку конкретной операции DOC-7781 и связанную проводку.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "drilldown_explain",
"actual_intent_class": "drilldown_explain",
"expected_route_hint": "live_mcp_drilldown",
"actual_route_hint": "live_mcp_drilldown",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": true,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": true,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "q6sYHRyO-swOjw",
"request_count_for_case": 1
},
{
"case_id": "NQ-004",
"raw_question": "По 97 счету проверь, где возможна ошибка дат начала и окончания списания.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "rule_based_account_control",
"actual_intent_class": "rule_based_account_control",
"expected_route_hint": "store_feature_risk",
"actual_route_hint": "store_feature_risk",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "fZskB6N3jC9x3h",
"request_count_for_case": 1
},
{
"case_id": "V11-RB-002",
"raw_question": "Проверь контрольные правила по ОС: где ошибки в сроках амортизации и учетной группе.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "rule_based_account_control",
"actual_intent_class": "rule_based_account_control",
"expected_route_hint": "store_feature_risk",
"actual_route_hint": "store_feature_risk",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "50lCHtVtBTYFXX",
"request_count_for_case": 1
},
{
"case_id": "V11-RB-003",
"raw_question": "По 10 счету проверь где нарушены правила оценки остатков.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "rule_based_account_control",
"actual_intent_class": "rule_based_account_control",
"expected_route_hint": "store_feature_risk",
"actual_route_hint": "store_feature_risk",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "xiLvYLKKIALKCI",
"request_count_for_case": 1
},
{
"case_id": "V11-RB-004",
"raw_question": "По НДС на 68.02 найди нарушения контрольных правил расчета за июнь.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "rule_based_account_control",
"actual_intent_class": "rule_based_account_control",
"expected_route_hint": "store_feature_risk",
"actual_route_hint": "store_feature_risk",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "oVnd_bUlvwAqko",
"request_count_for_case": 1
},
{
"case_id": "V11-RB-005",
"raw_question": "Проверь учетные настройки списания на 97 и покажи где высокий риск ручной ошибки.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "rule_based_account_control",
"actual_intent_class": "rule_based_account_control",
"expected_route_hint": "store_feature_risk",
"actual_route_hint": "store_feature_risk",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": true,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "DoZq1I-kFej9m0",
"request_count_for_case": 1
},
{
"case_id": "NQ-005",
"raw_question": "Есть ли аномальные материалы на счете 10, которые зависли и выглядят нелогично?",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "anomaly_probe",
"actual_intent_class": "anomaly_probe",
"expected_route_hint": "store_feature_risk",
"actual_route_hint": "store_feature_risk",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": true,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "3q9lkJMx_yJSym",
"request_count_for_case": 1
},
{
"case_id": "NQ-009",
"raw_question": "Где у нас пахнет ручной ошибкой по июню?",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "ambiguous_human_query",
"actual_intent_class": "ambiguous_human_query",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": false,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "VVeYqkgOUiyJUP",
"request_count_for_case": 1
},
{
"case_id": "V11-OT-003",
"raw_question": "Перед закрытием периода что у нас может взорваться в последний день?",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "period_close_risk",
"actual_intent_class": "period_close_risk",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "vYx_ndUvyOdrm1",
"request_count_for_case": 1
},
{
"case_id": "V11-OT-004",
"raw_question": "Где по июню выглядит подозрительно, но без точечного документа, просто дай зоны риска.",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "anomaly_probe",
"actual_intent_class": "anomaly_probe",
"expected_route_hint": "store_feature_risk",
"actual_route_hint": "store_feature_risk",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": false,
"needs_anomaly_summary": true,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "VRLcu2oMPHhfMP",
"request_count_for_case": 1
},
{
"case_id": "V11-OT-005",
"raw_question": "Че-то все криво на предзакрытии, где самые опасные места?",
"validation_passed": true,
"intent_match": true,
"route_match": true,
"causal_flags_match": true,
"expected_intent_class": "period_close_risk",
"actual_intent_class": "period_close_risk",
"expected_route_hint": "batch_refresh_then_store",
"actual_route_hint": "batch_refresh_then_store",
"expected_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false
},
"actual_requires": {
"needs_cross_entity_join": false,
"needs_causal_chain": false,
"needs_exact_object_trace": false,
"needs_ranking": true,
"needs_anomaly_summary": false,
"needs_runtime_truth": false,
"needs_period_cut": true,
"needs_evidence": false
},
"confidence_overall": "medium",
"trace_id": "0tXWviH21wzrlP",
"request_count_for_case": 1
}
]
}