feat(cost): persist reasoning tokens and the peak/off-peak tier

ob_ai_call is the only input to V2 pricing, and it was dropping the two
numbers needed to interpret its own cost column.

tokens_reasoning was captured from the provider response but never written to
the table, so the cost of thinking mode was invisible once it left the
process. Folding it into tokens_out was not an option either -- it is billed
at the output rate, and mixing the two hides exactly the difference we now
have to watch.

rate_tier records which price list applied. Without it a row's cost_cents
cannot be rechecked: DeepSeek's peak rate is double off-peak from 2026-08-17,
so recomputing an old row at "the current rate" silently produces a different
answer than what was actually billed.

Migration 0002 adds both columns. tokens_reasoning defaults to 0 rather than
NULL so cost aggregates do not collapse to NULL on historical rows;
rate_tier stays nullable because rows written before this change genuinely
have an unknown tier, and NULL says that more honestly than a guessed 'flat'.

The smoke script now asserts both: reasoning tokens must be 0 (a non-zero
value means a provider default re-enabled thinking mode behind our backs) and
rate_tier must be present. Verified against the real API -- tokens_reasoning=0,
rate_tier=off_peak, 3.9s round trip.

packages/llm/cost.py

@@ -31,6 +31,11 @@ PEAK_WINDOWS: tuple[tuple[time, time], ...] = (
     (time(14, 0), time(18, 0)),
 )
 
+# 记入 ob_ai_call.rate_tier，用于事后复核 cost_cents。
+TIER_PEAK = "peak"
+TIER_OFF_PEAK = "off_peak"
+TIER_FLAT = "flat"  # 只有一套单价（未启用峰谷，或该 provider 不分时段）
+
 
 @dataclass(frozen=True, slots=True)
 class ModelRate:
@@ -63,6 +68,9 @@ class RateTable:
     def known_models(self) -> tuple[str, ...]:
         return tuple(sorted(self._rates))
 
+    def current_tier(self, at: datetime | None = None) -> str:
+        return TIER_FLAT
+
 
 def is_peak(at: datetime | None = None) -> bool:
     """调用时刻是否落在供应商的高峰时段（按北京时间判定）。"""
@@ -94,6 +102,11 @@ class PeakAwareRateTable:
     def known_models(self) -> tuple[str, ...]:
         return self._off_peak.known_models()
 
+    def current_tier(self, at: datetime | None = None) -> str:
+        if not self._enabled:
+            return TIER_FLAT
+        return TIER_PEAK if is_peak(at) else TIER_OFF_PEAK
+
 
 def compute_cost_cents(usage: Usage, rate: ModelRate) -> Decimal:
     """Cost of one call, in cents, with cache-hit tokens billed separately."""
@@ -110,6 +123,10 @@ class RateResolver(Protocol):
 
     def for_model(self, model: str) -> ModelRate: ...
 
+    def current_tier(self) -> str:
+        """此刻适用的档位，随成本一起落库以便事后复核。"""
+        ...
+
 
 @dataclass(frozen=True, slots=True)
 class AICallRecord:
@@ -126,6 +143,7 @@ class AICallRecord:
     user_id: int | None = None
     prompt_id: str | None = None
     prompt_version: str | None = None
+    rate_tier: str | None = None
     error_code: str | None = None
     error_message: str | None = None
     ref_type: str | None = None