feat: update webui extensions, models, and agent config

packages/ai/src/image-models.generated.ts

@@ -125,6 +125,21 @@ export const IMAGE_MODELS = {
 				cacheWrite: 0,
 			},
 		} satisfies ImagesModel<"openrouter-images">,
+		"microsoft/mai-image-2.5": {
+			id: "microsoft/mai-image-2.5",
+			name: "Microsoft: MAI-Image-2.5",
+			api: "openrouter-images",
+			provider: "openrouter",
+			baseUrl: "https://openrouter.ai/api/v1",
+			input: ["text", "image"],
+			output: ["image"],
+			cost: {
+				input: 5,
+				output: 0,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+		} satisfies ImagesModel<"openrouter-images">,
 		"openai/gpt-5-image": {
 			id: "openai/gpt-5-image",
 			name: "OpenAI: GPT-5 Image",
@@ -425,6 +440,36 @@ export const IMAGE_MODELS = {
 				cacheWrite: 0,
 			},
 		} satisfies ImagesModel<"openrouter-images">,
+		"sourceful/riverflow-v2.5-fast": {
+			id: "sourceful/riverflow-v2.5-fast",
+			name: "Sourceful: Riverflow V2.5 Fast",
+			api: "openrouter-images",
+			provider: "openrouter",
+			baseUrl: "https://openrouter.ai/api/v1",
+			input: ["text", "image"],
+			output: ["image"],
+			cost: {
+				input: 0,
+				output: 0,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+		} satisfies ImagesModel<"openrouter-images">,
+		"sourceful/riverflow-v2.5-pro": {
+			id: "sourceful/riverflow-v2.5-pro",
+			name: "Sourceful: Riverflow V2.5 Pro",
+			api: "openrouter-images",
+			provider: "openrouter",
+			baseUrl: "https://openrouter.ai/api/v1",
+			input: ["text", "image"],
+			output: ["image"],
+			cost: {
+				input: 0,
+				output: 0,
+				cacheRead: 0,
+				cacheWrite: 0,
+			},
+		} satisfies ImagesModel<"openrouter-images">,
 		"x-ai/grok-imagine-image-quality": {
 			id: "x-ai/grok-imagine-image-quality",
 			name: "xAI: Grok Imagine Image Quality",

packages/ai/test/context-overflow.test.ts

@@ -126,9 +126,9 @@ describe("Context overflow error handling", () => {
 	describe("GitHub Copilot (OAuth)", () => {
 		// OpenAI model via Copilot
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should detect overflow via isContextOverflow",
+			"claude-sonnet-4.6 - should detect overflow via isContextOverflow",
 			async () => {
-				const model = getModel("github-copilot", "gpt-4o");
+				const model = getModel("github-copilot", "claude-sonnet-4.6");
 				const result = await testContextOverflow(model, githubCopilotToken!);
 				logResult(result);
 
@@ -297,8 +297,8 @@ describe("Context overflow error handling", () => {
 	// =============================================================================
 
 	describe.skipIf(!process.env.CEREBRAS_API_KEY)("Cerebras", () => {
-		it("qwen-3-235b - should detect overflow via isContextOverflow", async () => {
-			const model = getModel("cerebras", "qwen-3-235b-a22b-instruct-2507");
+		it("gpt-oss-120b - should detect overflow via isContextOverflow", async () => {
+			const model = getModel("cerebras", "gpt-oss-120b");
 			const result = await testContextOverflow(model, process.env.CEREBRAS_API_KEY!);
 			logResult(result);
 

packages/ai/test/empty.test.ts

@@ -630,37 +630,37 @@ describe("AI Providers Empty Message Tests", () => {
 
 	describe("GitHub Copilot Provider Empty Messages", () => {
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle empty content array",
+			"claude-sonnet-4.6 - should handle empty content array",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await testEmptyMessage(llm, { apiKey: githubCopilotToken });
 			},
 		);
 
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle empty string content",
+			"claude-sonnet-4.6 - should handle empty string content",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await testEmptyStringMessage(llm, { apiKey: githubCopilotToken });
 			},
 		);
 
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle whitespace-only content",
+			"claude-sonnet-4.6 - should handle whitespace-only content",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await testWhitespaceOnlyMessage(llm, { apiKey: githubCopilotToken });
 			},
 		);
 
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle empty assistant message in conversation",
+			"claude-sonnet-4.6 - should handle empty assistant message in conversation",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await testEmptyAssistantMessage(llm, { apiKey: githubCopilotToken });
 			},
 		);

packages/ai/test/fireworks-models.test.ts

@@ -38,7 +38,7 @@ describe("Fireworks models", () => {
 	});
 
 	it("registers the Fire Pass turbo router model", () => {
-		const model = getModel("fireworks", "accounts/fireworks/routers/kimi-k2p5-turbo");
+		const model = getModel("fireworks", "accounts/fireworks/routers/kimi-k2p6-turbo");
 
 		expect(model).toBeDefined();
 		expect(model.api).toBe("anthropic-messages");

packages/ai/test/image-tool-result.test.ts

@@ -443,19 +443,19 @@ describe("Tool Results with Images", () => {
 
 	describe("GitHub Copilot Provider", () => {
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle tool result with only image",
+			"claude-sonnet-4.6 - should handle tool result with only image",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await handleToolWithImageResult(llm, { apiKey: githubCopilotToken });
 			},
 		);
 
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle tool result with text and image",
+			"claude-sonnet-4.6 - should handle tool result with text and image",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await handleToolWithTextAndImageResult(llm, { apiKey: githubCopilotToken });
 			},
 		);

packages/ai/test/tokens.test.ts

@@ -149,7 +149,7 @@ describe("Token Statistics on Abort", () => {
 	});
 
 	describe.skipIf(!process.env.CEREBRAS_API_KEY)("Cerebras Provider", () => {
-		const llm = getModel("cerebras", "qwen-3-235b-a22b-instruct-2507");
+		const llm = getModel("cerebras", "gpt-oss-120b");
 
 		it("should include token stats when aborted mid-stream", { retry: 3, timeout: 30000 }, async () => {
 			await testTokensOnAbort(llm);
@@ -289,10 +289,10 @@ describe("Token Statistics on Abort", () => {
 
 	describe("GitHub Copilot Provider", () => {
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should include token stats when aborted mid-stream",
+			"claude-sonnet-4.6 - should include token stats when aborted mid-stream",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await testTokensOnAbort(llm, { apiKey: githubCopilotToken });
 			},
 		);

packages/ai/test/tool-call-without-result.test.ts

@@ -297,10 +297,10 @@ describe("Tool Call Without Result Tests", () => {
 
 	describe("GitHub Copilot Provider", () => {
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should filter out tool calls without corresponding tool results",
+			"claude-sonnet-4.6 - should filter out tool calls without corresponding tool results",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const model = getModel("github-copilot", "gpt-4o");
+				const model = getModel("github-copilot", "claude-sonnet-4.6");
 				await testToolCallWithoutResult(model, { apiKey: githubCopilotToken });
 			},
 		);

packages/ai/test/total-tokens.test.ts

@@ -666,10 +666,10 @@ describe("totalTokens field", () => {
 		);
 
 		it(
-			"google/gemini-2.0-flash-001 - should return totalTokens equal to sum of components",
+			"google/gemini-2.5-flash - should return totalTokens equal to sum of components",
 			{ retry: 3, timeout: 60000 },
 			async () => {
-				const llm = getModel("openrouter", "google/gemini-2.0-flash-001");
+				const llm = getModel("openrouter", "google/gemini-2.5-flash");
 
 				console.log(`\nOpenRouter / ${llm.id}:`);
 				const { first, second } = await testTotalTokensWithCache(llm, { apiKey: process.env.OPENROUTER_API_KEY });
@@ -706,10 +706,10 @@ describe("totalTokens field", () => {
 
 	describe("GitHub Copilot (OAuth)", () => {
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should return totalTokens equal to sum of components",
+			"claude-sonnet-4.6 - should return totalTokens equal to sum of components",
 			{ retry: 3, timeout: 60000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 
 				console.log(`\nGitHub Copilot / ${llm.id}:`);
 				const { first, second } = await testTotalTokensWithCache(llm, { apiKey: githubCopilotToken });

packages/ai/test/unicode-surrogate.test.ts

@@ -396,28 +396,28 @@ describe("AI Providers Unicode Surrogate Pair Tests", () => {
 
 	describe("GitHub Copilot Provider Unicode Handling", () => {
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle emoji in tool results",
+			"claude-sonnet-4.6 - should handle emoji in tool results",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await testEmojiInToolResults(llm, { apiKey: githubCopilotToken });
 			},
 		);
 
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle real-world LinkedIn comment data with emoji",
+			"claude-sonnet-4.6 - should handle real-world LinkedIn comment data with emoji",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await testRealWorldLinkedInData(llm, { apiKey: githubCopilotToken });
 			},
 		);
 
 		it.skipIf(!githubCopilotToken)(
-			"gpt-4o - should handle unpaired high surrogate (0xD83D) in tool results",
+			"claude-sonnet-4.6 - should handle unpaired high surrogate (0xD83D) in tool results",
 			{ retry: 3, timeout: 30000 },
 			async () => {
-				const llm = getModel("github-copilot", "gpt-4o");
+				const llm = getModel("github-copilot", "claude-sonnet-4.6");
 				await testUnpairedHighSurrogate(llm, { apiKey: githubCopilotToken });
 			},
 		);

packages/coding-agent/docs/rpc.md

@@ -426,7 +426,15 @@ Response:
 Execute a shell command and add output to conversation context.
 
 ```json
-{"type": "bash", "command": "ls -la"}
+{"type": "bash", "command": "ls -la", "excludeFromContext": false}
+```
+
+Optional `excludeFromContext` (default `false`): when `true`, the command output is shown in the UI but excluded from the LLM context on the next prompt (equivalent to `!!cmd` in the TUI).
+
+During execution, `bash_update` events stream partial stdout:
+
+```json
+{"type": "bash_update", "command": "ls -la", "partialOutput": "total 48\n"}
 ```
 
 Response:

packages/coding-agent/npm-shrinkwrap.json

@@ -1321,8 +1321,7 @@
 			"license": "MIT",
 			"bin": {
 				"jiti": "lib/jiti-cli.mjs"
-			},
-			"peer": true
+			}
 		},
 		"node_modules/json-bigint": {
 			"version": "1.0.0",
@@ -1775,8 +1774,7 @@
 			"license": "MIT",
 			"funding": {
 				"url": "https://github.com/sponsors/colinhacks"
-			},
-			"peer": true
+			}
 		},
 		"node_modules/zod-to-json-schema": {
 			"version": "3.25.2",

packages/coding-agent/src/modes/rpc/rpc-mode.ts

@@ -546,7 +546,15 @@ export async function runRpcMode(runtimeHost: AgentSessionRuntime): Promise<neve
 			// =================================================================
 
 			case "bash": {
-				const result = await session.executeBash(command.command);
+				let partialOutput = "";
+				const result = await session.executeBash(
+					command.command,
+					(chunk) => {
+						partialOutput += chunk;
+						output({ type: "bash_update", command: command.command, partialOutput });
+					},
+					{ excludeFromContext: command.excludeFromContext },
+				);
 				return success(id, "bash", result);
 			}
 

packages/coding-agent/src/modes/rpc/rpc-types.ts

@@ -50,7 +50,7 @@ export type RpcCommand =
 	| { id?: string; type: "abort_retry" }
 
 	// Bash
-	| { id?: string; type: "bash"; command: string }
+	| { id?: string; type: "bash"; command: string; excludeFromContext?: boolean }
 	| { id?: string; type: "abort_bash" }
 
 	// Session