Replace gr.Markdown thinking box with gr.HTML for reliable streaming

gr.Markdown visible toggling is unreliable in Gradio streaming generators.
Switched to gr.HTML with inline styles — empty string hides the element,
HTML string shows the styled box. No visibility state needed.

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

app.py

@@ -8,6 +8,7 @@
     YOULBOT_API_TOKEN=           ← api.py에 API_TOKEN 설정 시 동일 값
 """
 import asyncio
+import html as _html
 import os
 import platform
 import subprocess
@@ -95,14 +96,14 @@ async def tts_speak(text: str) -> str | None:
 
 async def respond(message, history, show_thinking, user_id, use_tts, run_ids):
     if not message.strip():
-        yield history, "", None, run_ids, gr.update()
+        yield history, "", None, run_ids, ""
         return
 
     history = list(history)
     run_ids = list(run_ids)
     history.append({"role": "user", "content": message})
     history.append({"role": "assistant", "content": ""})
-    yield history, "", None, run_ids, gr.update(value="", visible=False)
+    yield history, "", None, run_ids, ""  # thinking_box 초기화
 
     collected_run_id: str | None = None
     tts_text = ""            # 순수 답변만 누적 (TTS용)
@@ -118,29 +119,23 @@ async def respond(message, history, show_thinking, user_id, use_tts, run_ids):
             # 사고 과정(LLM thinking) — 박스에 추가
             if isinstance(token, dict) and "__thinking" in token:
                 thinking_acc += token["__thinking"]
-                yield history, "", None, run_ids, gr.update(
-                    value=f"🤔 **분석 중...**\n\n{thinking_acc}▌", visible=True
-                )
+                yield history, "", None, run_ids, _thinking_html(thinking_acc)
                 continue
 
             # 진행 로그(LangGraph, 검색 등) — 박스에 추가 (챗봇에는 표시 안 함)
             if isinstance(token, dict) and "__meta" in token:
                 thinking_acc += token["__meta"]
-                yield history, "", None, run_ids, gr.update(
-                    value=f"🤔 **분석 중...**\n\n{thinking_acc}▌", visible=True
-                )
+                yield history, "", None, run_ids, _thinking_html(thinking_acc)
                 continue
 
             # 첫 답변 토큰 도착 — 박스를 완료 상태로 전환
             if thinking_acc and not thinking_finalized:
                 thinking_finalized = True
-                yield history, "", None, run_ids, gr.update(
-                    value=f"💭 **분석 완료**\n\n{thinking_acc}", visible=True
-                )
+                yield history, "", None, run_ids, _thinking_html(thinking_acc, done=True)
 
             tts_text += token
             history[-1]["content"] += token
-            yield history, "", None, run_ids, gr.update()
+            yield history, "", None, run_ids, gr.update()  # thinking_box 유지
 
     except Exception as e:
         history[-1]["content"] += f"\n\n[오류: {e}]"
@@ -227,22 +222,25 @@ def delete_doc(source):
 
 # ── UI 구성 ──────────────────────────────────────────────────────
 
-_THINKING_CSS = """
-.thinking-box {
-    background: #f9f9f9;
-    border-left: 3px solid #bbb;
-    border-radius: 6px;
-    padding: 10px 14px;
-    margin-bottom: 6px;
-    max-height: 220px;
-    overflow-y: auto;
-    font-size: 0.85em;
-    color: #555;
-    white-space: pre-wrap;
-}
-"""
+_THINKING_STYLE = (
+    "background:#f9f9f9;border-left:3px solid #bbb;border-radius:6px;"
+    "padding:10px 14px;max-height:220px;overflow-y:auto;"
+    "font-size:0.85em;color:#555;white-space:pre-wrap;margin-bottom:6px;"
+)
 
-with gr.Blocks(title="율봇", css=_THINKING_CSS) as demo:
+
+def _thinking_html(text: str, done: bool = False) -> str:
+    icon = "💭" if done else "🤔"
+    label = "분석 완료" if done else "분석 중..."
+    cursor = "" if done else " ▌"
+    return (
+        f'<div style="{_THINKING_STYLE}">'
+        f"<strong>{icon} {label}</strong><br><br>"
+        f"{_html.escape(text)}{cursor}</div>"
+    )
+
+
+with gr.Blocks(title="율봇") as demo:
     gr.Markdown("# 율봇\n육아·금융 전문 AI 상담 도우미")
 
     user_state = gr.State(DEFAULT_USER)
@@ -257,11 +255,7 @@ with gr.Blocks(title="율봇", css=_THINKING_CSS) as demo:
                 scale=1,
             )
 
-        thinking_box = gr.Markdown(
-            value="",
-            visible=False,
-            elem_classes=["thinking-box"],
-        )
+        thinking_box = gr.HTML(value="")
         chatbot = gr.Chatbot(label="율봇", height=500)
         with gr.Row():
             msg_box = gr.Textbox(