refactor(api): TOML 配置 SSOT、统一错误契约、Auth/事务加固与可观测性 (#33)

配置 SSOT（TOML + .env）
统一错误契约
Auth 与事务边界
Redis / Celery 可靠性:业务 Redis（DB/0）与 Celery broker/backend（DB/1）显式拆分；连接池、sync client
可观测性（OpenTelemetry + LGTM）

api/tests/test_chat_input_normalize.py

@@ -2,6 +2,7 @@
 
 from unittest.mock import MagicMock, patch
 
+from app.features.conversation.constants import chat
 from app.features.conversation.input_normalize import (
     apply_conversation_input_rules,
     normalize_chat_input_for_agent,
@@ -15,39 +16,39 @@ def test_apply_conversation_rules_matches_memoir_mei_kanshang() -> None:
 
 def test_normalize_chat_rules_mode() -> None:
     raw = "美看上我"
-    with patch("app.features.conversation.input_normalize.settings") as m:
-        m.chat_input_normalize_enabled = True
-        m.chat_input_normalize_mode = "rules"
-        m.chat_input_normalize_llm_max_tokens = 512
-        m.chat_input_normalize_llm_max_input_chars = 8000
+    with patch("app.features.conversation.input_normalize.chat") as m:
+        m.input_normalize_enabled = True
+        m.input_normalize_mode = "rules"
+        m.input_normalize_llm_max_tokens = 512
+        m.input_normalize_llm_max_input_chars = 8000
         assert normalize_chat_input_for_agent(raw, llm=None) == "没看上我"
 
 
 def test_normalize_chat_disabled_returns_raw() -> None:
     raw = "美看上我"
-    with patch("app.features.conversation.input_normalize.settings") as m:
-        m.chat_input_normalize_enabled = False
-        m.chat_input_normalize_mode = "rules"
+    with patch("app.features.conversation.input_normalize.chat") as m:
+        m.input_normalize_enabled = False
+        m.input_normalize_mode = "rules"
         assert normalize_chat_input_for_agent(raw, llm=None) == raw
 
 
 def test_normalize_chat_off_mode() -> None:
     raw = "美看上我"
-    with patch("app.features.conversation.input_normalize.settings") as m:
-        m.chat_input_normalize_enabled = True
-        m.chat_input_normalize_mode = "off"
+    with patch("app.features.conversation.input_normalize.chat") as m:
+        m.input_normalize_enabled = True
+        m.input_normalize_mode = "off"
         assert normalize_chat_input_for_agent(raw, llm=None) == raw
 
 
 def test_normalize_llm_mode_voice_only_passes_no_llm_for_typing() -> None:
     raw = "美看上我"
     fake = MagicMock()
-    with patch("app.features.conversation.input_normalize.settings") as m:
-        m.chat_input_normalize_enabled = True
-        m.chat_input_normalize_mode = "llm"
-        m.chat_input_normalize_llm_voice_only = True
-        m.chat_input_normalize_llm_max_tokens = 512
-        m.chat_input_normalize_llm_max_input_chars = 8000
+    with patch("app.features.conversation.input_normalize.chat") as m:
+        m.input_normalize_enabled = True
+        m.input_normalize_mode = "llm"
+        m.input_normalize_llm_voice_only = True
+        m.input_normalize_llm_max_tokens = 512
+        m.input_normalize_llm_max_input_chars = 8000
         with patch(
             "app.features.conversation.input_normalize._llm_normalize_chat_input"
         ) as llm_norm: