test(ui): derive Anthropic Family reasoning model from preset

The 'carries a preset's per-tier reasoning effort through to the create
payload' assertion hardcoded claude-opus-5, which broke after #40341
retargeted the Anthropic Family reasoning tier at claude-fable-5-1. Read
the model from ANTHROPIC_TIERS so future preset edits don't red this
test, matching the pattern the file already documents at ALL_FAMILY_MODELS.

Co-authored-by: Krrish Dholakia <krrish-berri-2@users.noreply.github.com>

ui/litellm-dashboard/src/components/add_model/add_auto_router_tab.test.tsx

@@ -1047,7 +1047,7 @@ describe("AddAutoRouterTab", () => {
       expect(vi.mocked(handleAddAutoRouterSubmit).mock.calls.at(-1)?.[0]).toMatchObject({
         complexity_router_config: {
           tier_model_configs: {
-            REASONING: [{ model_name: "claude-opus-5", litellm_params: { reasoning_effort: "high" } }],
+            REASONING: [{ model_name: ANTHROPIC_TIERS.REASONING[0], litellm_params: { reasoning_effort: "high" } }],
           },
         },
       });