fix: revive the reasoning verdict as an enum, not a bare string

asdict keeps the enum and json.dumps writes it as a string because
StrEnum is a str subclass, but nothing turns it back on the way in, so
a cache hit returned a plain str where the annotation promised an enum.
Verified end to end rather than assumed from the subclass relation.

A value outside the domain now raises inside the existing guard and the
call falls back to source, which is the right direction for a poisoned
or stale cache entry. Entries written before this column existed still
replay: the guard checks for the key first, and a test pins that, since
turning it into an unconditional conversion would quietly turn every
pre-upgrade entry into a permanent miss.

src/polygateway/middleware/cache.py

@@ -17,7 +17,7 @@ from typing import TYPE_CHECKING, Any
 
 from loguru import logger
 
-from polygateway.types import ChatRequest, LLMResponse
+from polygateway.types import ChatRequest, LLMResponse, ThinkingObservation
 
 if TYPE_CHECKING:
     from collections.abc import Mapping
@@ -132,6 +132,12 @@ class CacheMW:
             data = json.loads(raw)
             fields = {k: v for k, v in data.items() if k in _RESPONSE_FIELDS}
             structured_data = self._rebuild_structured(fields.get("content", ""), request)
+            # JSON 里存的是 StrEnum 的字符串值,不转就复活成裸 str,与字段注解分叉
+            # (下游 `is ThinkingObservation.OBSERVED` 会在命中路径上静默为 False);
+            # 键缺失即升级前写入的旧条目,交给 dataclass 默认值。域外值抛
+            # ValueError,由下方 except 吞成"按未命中回源"——降级方向正确。
+            if "thinking_observation" in fields:
+                fields["thinking_observation"] = ThinkingObservation(fields["thinking_observation"])
             fields.update(
                 cache_hit=True,
                 latency_ms=0,