feat(quick-pane): persist default thinking level

ui/src/App.test.tsx

@@ -147,6 +147,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "gpt-5",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		mocks.invoke.mockImplementation((command: string) => {
@@ -189,6 +190,10 @@ describe("App command progress and errors", () => {
 		await user.click(
 			screen.getByRole("option", { name: /Claude.*anthropic\/claude\/sonnet/i }),
 		);
+		await user.selectOptions(
+			screen.getByRole("combobox", { name: "Default thinking" }),
+			"high",
+		);
 		await user.click(
 			screen.getByRole("button", { name: "Save Quick Pane settings" }),
 		);
@@ -200,6 +205,7 @@ describe("App command progress and errors", () => {
 						...loaded.quickPane,
 						provider: "openrouter",
 						model: "anthropic/claude/sonnet",
+						defaultThinkingLevel: "high",
 					},
 				},
 			}),
@@ -222,6 +228,7 @@ describe("App command progress and errors", () => {
 							followUpModel: "old-model",
 							instructions: "Stay concise.",
 							maximumToolCalls: 6,
+							defaultThinkingLevel: "medium",
 						},
 					});
 				if (command === "create_quick_runtime") {
@@ -283,6 +290,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "old-model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		const first = deferred<{
@@ -357,6 +365,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "old-model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		const staleSnapshot = deferred<{
@@ -450,6 +459,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		const snapshot = deferred<unknown>();
@@ -487,6 +497,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "old-model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		let openCount = 0;
@@ -546,6 +557,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "old-model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		let runtimeCount = 0;
@@ -620,6 +632,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "old-model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		const oldSnapshot = deferred<{
@@ -693,6 +706,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		let runtimeCount = 0;
@@ -764,6 +778,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		let createCount = 0;
@@ -819,6 +834,7 @@ describe("App command progress and errors", () => {
 				followUpModel: "model",
 				instructions: "Stay concise.",
 				maximumToolCalls: 6,
+				defaultThinkingLevel: "medium",
 			},
 		};
 		const nextSnapshot = deferred<unknown>();