修复测试模型路径并完善四路八路回归与文档

- 按文件哈希区分当前八路工程和历史 Amesim 基准，修复测试与清单中的旧路径。
- 四路审计从 JSON 按需生成 XML，八路新增只核对模式，并显式使用 UTF-8 与 LF。
- 将缺失高刚度夹具的测试替换为当前八路完整 BDF 运行，明确未恢复原 RK45 专项覆盖。
- 更新前端大型工程测试、历史活动回放入口、数据说明、现行规范和相关报告。

验证：全量后台 374 项通过、1 项条件跳过；前端 11 项通过、历史独立服务用例 1 项跳过；四路八路 AME 审计通过。
环境、缓存和运行产物保持忽略；本次未实施工程 JSON 字段精简。

tools/audit_test_mql8_model.py

@@ -1,11 +1,14 @@
 """Reproduce the eight-branch AME drawing -> editor model audit.
 
-Run from the repository root with .venv/bin/python. The original inputs are
-read only. Mapping uses typed graph isomorphism and endpoint edit cost, never
-component display names; names are emitted only as evidence for review.
+Run from the repository root with .venv/bin/python and --check to preserve
+the current project; omit --check to regenerate it. Audit evidence goes to
+the ignored test/model-audit/mql8 directory. Mapping uses typed graph
+isomorphism and endpoint edit cost, never component display names; names
+are emitted only as evidence for review.
 """
 from __future__ import annotations
 
+import argparse
 from collections import Counter, defaultdict
 from copy import deepcopy
 import hashlib
@@ -24,11 +27,12 @@ from tests.test_test_mql_ame_contract import (
     _SOURCE_FIELD_BY_PARAMETER, _numeric_project_value,
 )
 from app.simulation.registry import get_component_model_spec
+from app.simulation.paths import CURRENT_MQL8_AME_PATH, CURRENT_MQL8_PROJECT_PATH
 
-HERE = ROOT / 'test/solver-newton-20260911/mql8'
-AME = ROOT / 'tests/data/test_mql.ame'
-INPUT = ROOT / 'tests/fixtures/legacy/test-mql-8.json'
-OUTPUT = ROOT / 'tests/data/test-mql-8-corrected.json'
+HERE = ROOT / 'test/model-audit/mql8'
+AME = CURRENT_MQL8_AME_PATH
+INPUT = CURRENT_MQL8_PROJECT_PATH
+OUTPUT = CURRENT_MQL8_PROJECT_PATH
 
 
 def read_drawing():
@@ -179,8 +183,14 @@ def match_graph(entities, ame_edges, project):
 
 
 def main():
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument('--check', action='store_true',
+                        help='Audit current JSON against AME without changing the project; evidence goes to test/model-audit/mql8/.')
+    arguments = parser.parse_args()
     HERE.mkdir(parents=True, exist_ok=True)
-    original = json.loads(INPUT.read_text())
+    sources = {str(p.relative_to(ROOT)): {'sha256': hashlib.sha256(p.read_bytes()).hexdigest(),
+               'bytes': p.stat().st_size} for p in (AME, INPUT)}
+    original = json.loads(INPUT.read_text(encoding='utf-8'))
     entities, edges, globals_, sim, model_info = read_drawing()
     mapping, matching = match_graph(entities, edges, original)
     corrected = deepcopy(original)
@@ -266,7 +276,8 @@ def main():
     payload = ReactFlowProjectPayload.model_validate(executable)
     network = compile_reactflow_network(payload)
     program = compile_native_program(network)
-    OUTPUT.write_text(json.dumps(corrected, ensure_ascii=False, indent=2) + '\n')
+    if not arguments.check:
+        OUTPUT.write_text(json.dumps(corrected, ensure_ascii=False, indent=2) + '\n', encoding='utf-8', newline='\n')
     xml = build_reactflow_system_xml(payload)
     (HERE / 'corrected.xml').write_bytes(xml if isinstance(xml, bytes) else xml.encode())
     # The archive's parameter table and run input have matching row order.
@@ -296,7 +307,8 @@ def main():
         cache_checks.append({'dataPath': path, 'parameterDataLine': line,
                              'drawingValue': drawing_value, 'runValue': run_value})
     audit = {
-        'sources': {str(p.relative_to(ROOT)): {'sha256': hashlib.sha256(p.read_bytes()).hexdigest(), 'bytes': p.stat().st_size} for p in (AME, INPUT)},
+        'sources': sources,
+        'checkOnly': arguments.check,
         'output': str(OUTPUT.relative_to(ROOT)), 'globals': globals_, 'matching': matching,
         'counts': {'components': len(entities), 'connections': len(edges), 'parameters': len(parameter_audit),
                    'changedParameters': len(parameter_changes), 'changedConnections': len(connection_changes),
@@ -315,11 +327,16 @@ def main():
         'parameters': parameter_audit, 'connections': connection_audit,
         'evaluationSchedule': program.evaluation_schedule,
     }
-    (HERE / 'audit.json').write_text(json.dumps(audit, ensure_ascii=False, indent=2) + '\n')
+    (HERE / 'audit.json').write_text(json.dumps(audit, ensure_ascii=False, indent=2) + '\n', encoding='utf-8', newline='\n')
     print(json.dumps({'counts': audit['counts'], 'matching': matching,
                       'parameters': parameter_changes, 'connections': connection_changes,
                       'simulation': audit['simulation']}, ensure_ascii=False, indent=2))
 
+    if arguments.check:
+        assert not parameter_changes, 'Current JSON parameters differ from AME; see audit.json.'
+        assert not connection_changes, 'Current JSON connection ports differ from AME; see audit.json.'
+        assert original['simulation'] == corrected['simulation'], 'Current JSON time settings differ from AME.'
+
 
 if __name__ == '__main__':
     main()