feat: Phase 2 complete — 13 Phases of COBOL type classification and test benchmark

P0.6: gcov infrastructure
P1: extract_structure output expansion (11 new feature fields)
P2: Confusion group rule engine (8 pairs + contradiction + backtrack)
P3: 4-factor confidence calculation + quality gate update
P4: 33+2 COBOL program type test samples (22 files, 7 categories)
P5: parametrized/ test data generation engine
P6: japanese_data.py lookup tables
P7-10: Type-specific test suites (~159 parametrized tests)
P11: Full classification pipeline (classify_program) + orchestrator integration
P12: Documentation (module-interfaces, test-plan v3.0, coverage-matrix)

Architecture decisions:
- classification_pipeline/ merged to hina/pipeline/
- parametrized/ as independent module
- japanese_data.py as root-level file
- hina/__all__ only exports classify_program()

Co-Authored-By: Claude <noreply@anthropic.com>

test_llm.py

@@ -0,0 +1,17 @@
+import json, os, sys
+sys.path.insert(0, ".")
+
+os.environ["LLM_API_KEY"] = "sk-ca4961087c7f4aefa8ed0fc6f3d02329"
+os.environ["LLM_API_BASE"] = "https://api.deepseek.com/v1"
+
+from agents.llm import LLMClient
+import time
+
+c = LLMClient(model="deepseek-chat", timeout=30)
+t0 = time.time()
+r = c.call([
+    {"role":"system","content":"Parse this COBOL COPYBOOK into JSON: {\"fields\":[{\"name\":\"...\",\"level\":N,\"pic\":\"...\",\"usage\":\"DISPLAY|COMP-3\",\"length\":N}]}"},
+    {"role":"user","content": open("uploads/ec17bf32/copybook.cpy").read()}
+])
+print(f"LLM call OK ({time.time()-t0:.1f}s)")
+print(r[:500])