ai-labeling: Prompt 1 v0.1 + ground-truth v0.4.0(20期) + MiMo首轮跑批

- Prompt 1 v0.1: 4分类字段(题材/装备域/场景/技术)判别prompt，含5示例+边界规则+换装测试
- ground-truth v0.4.0: 10→20期扩展，ep编号重映射(旧ep3-15→新ep001-020)，4分类字段全填(制片人逐期审定)
- 文稿升级: 旧10期删除，新20期md文件(doco子项目清洗产出)替换
- 脚本升级: ep编号2位→3位，ALL_EPISODES扩至1-20，summarize classification比对逻辑实现
- MiMo首轮结果: 题材75%/装备域70%/场景95%/技术70%/全对40%，需迭代Prompt
- 质量标杆: 加入PPT+Excel样板文件(example/)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

ai-labeling/scripts/run_labeling.py

@@ -39,7 +39,7 @@ MODEL_CONFIG = {
     },
 }
 
-ALL_EPISODES = [3, 4, 5, 7, 10, 11, 12, 13, 14, 15]
+ALL_EPISODES = list(range(1, 21))
 
 FIELD_PROMPT_MAP = {
     "narrative": "prompt2_narrative.md",
@@ -54,7 +54,7 @@ def load_prompt(field):
 
 
 def load_transcript(ep):
-    pattern = f"ep{ep:02d}_*.md"
+    pattern = f"ep{ep:03d}_*.md"
     files = list(TRANSCRIPTS_DIR.glob(pattern))
     if not files:
         raise FileNotFoundError(f"No transcript found for ep{ep:02d} in {TRANSCRIPTS_DIR}")
@@ -131,12 +131,12 @@ def run_labeling(ep, model_key, field="narrative"):
     result = call_model(model_key, system_prompt, user_prompt)
     gt = load_ground_truth(ep)
     ts = datetime.now().strftime("%Y%m%d_%H%M%S")
-    out = EXPERIMENTS_DIR / f"{ts}_{model_key}_{field}_ep{ep:02d}.json"
+    out = EXPERIMENTS_DIR / f"{ts}_{model_key}_{field}_ep{ep:03d}.json"
     out.write_text(
         json.dumps({"episode": ep, "filename": fname, "field": field, "result": result, "ground_truth": gt}, ensure_ascii=False, indent=2),
         encoding="utf-8",
     )
-    print(f"完成 ep{ep:02d} [{field}] -> {out.name}")
+    print(f"完成 ep{ep:03d} [{field}] -> {out.name}")
     return result