feat(llm): add Claude Sonnet 5 and Fable 5 to the catalog

- refresh the Anthropic lineup to the current GA set (Fable 5, Opus 4.8,
  Sonnet 5, Opus 4.7, Haiku 4.5)
- extend the effort gate to single-number Claude 5 IDs (claude-sonnet-5,
  claude-fable-5) so their effort setting is honored

tests/test_anthropic_effort.py

@@ -57,6 +57,16 @@ class TestEffortGate:
         mod.AnthropicClient(model=model, effort="low", api_key="x").get_llm()
         assert captured["kwargs"]["effort"] == "low"
 
+    @pytest.mark.parametrize(
+        "model",
+        # Claude 5 family uses single-number version IDs; all are effort-capable.
+        ["claude-sonnet-5", "claude-fable-5", "claude-mythos-5"],
+    )
+    def test_claude_5_family_receives_effort(self, monkeypatch, model):
+        captured = _capture_kwargs(monkeypatch)
+        mod.AnthropicClient(model=model, effort="high", api_key="x").get_llm()
+        assert captured["kwargs"]["effort"] == "high"
+
     def test_mythos_preview_receives_effort(self, monkeypatch):
         captured = _capture_kwargs(monkeypatch)
         mod.AnthropicClient(

tests/test_temperature_config.py

@@ -16,9 +16,12 @@ class TestTemperatureForwarding:
     @pytest.mark.parametrize(
         "provider,model",
         [
+            # gpt-4.1 is intentionally a non-reasoning model: the GPT-5 family
+            # are reasoning models and correctly drop temperature (see
+            # test_openai_reasoning_effort), so forwarding is tested on gpt-4.1.
             ("openai", "gpt-4.1"),
-            ("anthropic", "claude-sonnet-4-6"),
-            ("google", "gemini-2.5-flash"),
+            ("anthropic", "claude-sonnet-5"),
+            ("google", "gemini-3.5-flash"),
             ("deepseek", "deepseek-chat"),
         ],
     )