fix: count gate INFRA guard numerator by unit not record

app/harness/validate.py

@@ -259,6 +259,23 @@ def _infra_question_ids_from_db(
     }
 
 
+def _count_infra_units(units: list[QuestionUnit], infra_qids: set[str]) -> int:
+    """统计含 INFRA record 的 unit 数（一个 unit 任一题 INFRA 即计 1）。
+
+    使护栏分子与分母（r.total，unit 粒度）同口径：AR pair 一 unit 含两 record，
+    逐 record 计数会放大分子致 gate_guard_err 误触发，破坏 unit 粒度一致性
+    （核心算法保真 #5/#6）。
+
+    参数:
+        units: 当前块的单元列表（single 或 AR pair）。
+        infra_qids: 本 run 中 stop_reason 属 INFRA 故障族的 question_id 集合。
+
+    返回:
+        含至少一题 INFRA 的 unit 数。
+    """
+    return sum(1 for u in units if any(q.question_id in infra_qids for q in u.questions))
+
+
 def _candidate_correctness_from_db(
     log: HarnessLog,
     run_id: str,
@@ -332,11 +349,12 @@ async def _resolve_baseline_block(
     if miss_units:
         miss_questions = flatten_units(miss_units)
         r_b = await run_inference(miss_questions, run_id=run_id, skills_dir=base_skills_dir)
-        # 护栏错误计数与 INFRA 判定口径一致：error + parse_error 都计入，
-        # 使 parse_error 风暴同样能触发 gate_guard_err 熔断（不被绕过）。
-        errors_inc = sum(r_b.stop_reason_counts.get(reason, 0) for reason in _INFRA_STOP_REASONS)
-        denom_inc = r_b.total
         infra_qids = _infra_question_ids_from_db(log, r_b.run_id, miss_questions)
+        # 护栏分子与分母（r.total，unit 粒度）同口径：含 INFRA record 的 unit 计 1，
+        # 避免 AR pair（一 unit 两 record）逐 record 计数放大分子致误触发；仍涵盖
+        # error + parse_error（_infra_question_ids_from_db 口径），parse_error 风暴不被绕过。
+        errors_inc = _count_infra_units(miss_units, infra_qids)
+        denom_inc = r_b.total
         fresh_per_q = _candidate_correctness_from_db(log, r_b.run_id, miss_questions)
         fresh_units = unit_correctness_view(miss_units, fresh_per_q)
         # 只回写非 INFRA 单元；INFRA 单元不入缓存（不永久污染基线快照）
@@ -382,8 +400,10 @@ async def _run_candidate_block(
     questions = flatten_units(units)
     r_c = await run_inference(questions, run_id=run_id, skills_dir=cand_dir)
     c_per_q = _candidate_correctness_from_db(log, r_c.run_id, questions)
-    # 护栏错误计数与 INFRA 判定口径一致：error + parse_error 都计入。
-    errors_inc = sum(r_c.stop_reason_counts.get(reason, 0) for reason in _INFRA_STOP_REASONS)
+    infra_qids = _infra_question_ids_from_db(log, r_c.run_id, questions)
+    # 护栏分子与分母（r.total，unit 粒度）同口径：含 INFRA record 的 unit 计 1
+    # （见 _count_infra_units），涵盖 error + parse_error。
+    errors_inc = _count_infra_units(units, infra_qids)
     return c_per_q, errors_inc, r_c.total