feat: preserve governed OpenAI tool messages in 1.4.0

Add faithful single-choice chat results, tool stream activity, error-frame retries, and isolated protocol caching.

Reject empty cached completions with the same value predicate as transport, add regression coverage, and adopt Apache-2.0.

tests/unit/test_config.py

@@ -1397,9 +1397,7 @@ class TestStallConfig:
     def test_stall_after_s_single_source_warns(self):
         """单源 scope 设 stall 阈值: 装配期 warning 放行(1.3.7 §5 同款)。"""
         with _captured_warnings() as warnings:
-            s = GatewaySettings.from_env(
-                "LLM", env=_env(**{"LLM__HEDGE__STALL_AFTER_S": "8"})
-            )
+            s = GatewaySettings.from_env("LLM", env=_env(**{"LLM__HEDGE__STALL_AFTER_S": "8"}))
         assert s.hedge_stall_after_s == 8.0
         assert any("单源" in m for m in warnings)
 
@@ -1422,9 +1420,7 @@ class TestStallConfig:
             "EMBED__BATCH_SIZE": "64",
         }
         with _captured_warnings() as warnings:
-            EmbeddingSettings.from_env(
-                "EMBED", env={**embed_env, "EMBED__HEDGE__AFTER_S": "8"}
-            )
+            EmbeddingSettings.from_env("EMBED", env={**embed_env, "EMBED__HEDGE__AFTER_S": "8"})
         assert any("仅对 chat scope 生效" in m for m in warnings)
         with _captured_warnings() as warnings:
             EmbeddingSettings.from_env(