feat: optimize todo_tool with multi-tool execution and diff UI

- Enable update_todo_list to execute alongside other tools in the same LLM response
- Replace full todo list output with collapsible diff-style UI feedback
- Add TodoListDiff component for compact change visualization
- Implement diff generation algorithm to track added/removed/modified todos
- Add comprehensive tests for multi-tool execution and diff generation
- Reduce token usage by showing only changes instead of full list

Fixes #7412

Author: roomote

src/core/assistant-message/__tests__/presentAssistantMessage.multi-tool.spec.ts

@@ -0,0 +1,199 @@
+import { describe, it, expect, vi, beforeEach } from "vitest"
+import { presentAssistantMessage } from "../presentAssistantMessage"
+import { Task } from "../../task/Task"
+import { updateTodoListTool } from "../../tools/updateTodoListTool"
+import { readFileTool } from "../../tools/readFileTool"
+
+vi.mock("../../tools/updateTodoListTool")
+vi.mock("../../tools/readFileTool")
+vi.mock("@roo-code/telemetry", () => ({
+	TelemetryService: {
+		instance: {
+			captureToolUsage: vi.fn(),
+			captureConsecutiveMistakeError: vi.fn(),
+		},
+		hasInstance: () => true,
+	},
+}))
+
+describe("presentAssistantMessage - Multi-tool execution", () => {
+	let mockTask: Task
+	let mockAskApproval: any
+	let mockHandleError: any
+	let mockPushToolResult: any
+
+	beforeEach(() => {
+		vi.clearAllMocks()
+
+		// Create a minimal mock task
+		mockTask = {
+			taskId: "test-task",
+			instanceId: "test-instance",
+			abort: false,
+			presentAssistantMessageLocked: false,
+			presentAssistantMessageHasPendingUpdates: false,
+			currentStreamingContentIndex: 0,
+			assistantMessageContent: [],
+			didCompleteReadingStream: false,
+			userMessageContentReady: false,
+			didRejectTool: false,
+			didAlreadyUseTool: false,
+			userMessageContent: [],
+			say: vi.fn(),
+			ask: vi.fn(),
+			recordToolUsage: vi.fn(),
+			recordToolError: vi.fn(),
+			consecutiveMistakeCount: 0,
+			clineMessages: [],
+			apiConversationHistory: [],
+			todoList: [],
+			checkpointSave: vi.fn(),
+			currentStreamingDidCheckpoint: false,
+			browserSession: { closeBrowser: vi.fn() },
+			toolRepetitionDetector: { check: vi.fn(() => ({ allowExecution: true })) },
+			providerRef: { deref: vi.fn(() => ({ getState: vi.fn(() => ({ mode: "code" })) })) },
+			api: { getModel: vi.fn(() => ({ id: "test-model" })) },
+		} as any
+
+		mockAskApproval = vi.fn(() => Promise.resolve(true))
+		mockHandleError = vi.fn()
+		mockPushToolResult = vi.fn()
+	})
+
+	it("should allow update_todo_list to execute alongside other tools", async () => {
+		// Set up assistant message content with two tools
+		mockTask.assistantMessageContent = [
+			{
+				type: "tool_use",
+				name: "read_file",
+				params: { path: "test.txt" },
+				partial: false,
+			},
+			{
+				type: "tool_use",
+				name: "update_todo_list",
+				params: { todos: "[ ] Test todo" },
+				partial: false,
+			},
+		]
+
+		// Mock the tool implementations
+		vi.mocked(readFileTool).mockImplementation(async (cline, block, askApproval, handleError, pushToolResult) => {
+			pushToolResult("File content")
+		})
+
+		vi.mocked(updateTodoListTool).mockImplementation(
+			async (cline, block, askApproval, handleError, pushToolResult) => {
+				pushToolResult("Todo list updated")
+			},
+		)
+
+		// Process first tool
+		mockTask.currentStreamingContentIndex = 0
+		await presentAssistantMessage(mockTask)
+
+		// After first tool, didAlreadyUseTool should be true
+		expect(mockTask.didAlreadyUseTool).toBe(true)
+
+		// Process second tool (update_todo_list)
+		mockTask.currentStreamingContentIndex = 1
+		await presentAssistantMessage(mockTask)
+
+		// Both tools should have been executed
+		expect(readFileTool).toHaveBeenCalledTimes(1)
+		expect(updateTodoListTool).toHaveBeenCalledTimes(1)
+
+		// Check that both tool results were pushed
+		// The first two entries should be for read_file
+		expect(mockTask.userMessageContent[0]).toEqual({
+			type: "text",
+			text: expect.stringContaining("Result:"),
+		})
+		expect(mockTask.userMessageContent[1]).toEqual({
+			type: "text",
+			text: "File content",
+		})
+		// The next two entries should be for update_todo_list
+		expect(mockTask.userMessageContent[2]).toEqual({
+			type: "text",
+			text: "[update_todo_list] Result:",
+		})
+		expect(mockTask.userMessageContent[3]).toEqual({
+			type: "text",
+			text: "Todo list updated",
+		})
+	})
+
+	it("should block non-update_todo_list tools after a tool has been used", async () => {
+		// Set up assistant message content with two non-update_todo_list tools
+		mockTask.assistantMessageContent = [
+			{
+				type: "tool_use",
+				name: "read_file",
+				params: { path: "test.txt" },
+				partial: false,
+			},
+			{
+				type: "tool_use",
+				name: "write_to_file",
+				params: { path: "test.txt", content: "new content" },
+				partial: false,
+			},
+		]
+
+		// Mock the read_file tool
+		vi.mocked(readFileTool).mockImplementation(async (cline, block, askApproval, handleError, pushToolResult) => {
+			pushToolResult("File content")
+		})
+
+		// Process first tool
+		mockTask.currentStreamingContentIndex = 0
+		await presentAssistantMessage(mockTask)
+
+		// After first tool, didAlreadyUseTool should be true
+		expect(mockTask.didAlreadyUseTool).toBe(true)
+
+		// Process second tool (should be blocked)
+		mockTask.currentStreamingContentIndex = 1
+		await presentAssistantMessage(mockTask)
+
+		// Only the first tool should have been executed
+		expect(readFileTool).toHaveBeenCalledTimes(1)
+
+		// Check that the second tool was blocked
+		expect(mockTask.userMessageContent).toContainEqual(
+			expect.objectContaining({
+				type: "text",
+				text: expect.stringContaining(
+					"Tool [write_to_file] was not executed because a tool has already been used",
+				),
+			}),
+		)
+	})
+
+	it("should not set didAlreadyUseTool when update_todo_list is executed", async () => {
+		// Set up assistant message content with update_todo_list first
+		mockTask.assistantMessageContent = [
+			{
+				type: "tool_use",
+				name: "update_todo_list",
+				params: { todos: "[ ] Test todo" },
+				partial: false,
+			},
+		]
+
+		// Mock the tool implementation
+		vi.mocked(updateTodoListTool).mockImplementation(
+			async (cline, block, askApproval, handleError, pushToolResult) => {
+				pushToolResult("Todo list updated")
+			},
+		)
+
+		// Process the update_todo_list tool
+		await presentAssistantMessage(mockTask)
+
+		// didAlreadyUseTool should remain false for update_todo_list
+		expect(mockTask.didAlreadyUseTool).toBe(false)
+		expect(updateTodoListTool).toHaveBeenCalledTimes(1)
+	})
+})

src/core/assistant-message/presentAssistantMessage.ts

@@ -242,8 +242,11 @@ export async function presentAssistantMessage(cline: Task) {
 				break
 			}
 
-			if (cline.didAlreadyUseTool) {
-				// Ignore any content after a tool has already been used.
+			// Special handling for update_todo_list - it can be used alongside other tools
+			const isUpdateTodoList = block.name === "update_todo_list"
+
+			if (cline.didAlreadyUseTool && !isUpdateTodoList) {
+				// Ignore any content after a tool has already been used (except update_todo_list).
 				cline.userMessageContent.push({
 					type: "text",
 					text: `Tool [${block.name}] was not executed because a tool has already been used in this message. Only one tool may be used per message. You must assess the first tool's result before proceeding to use the next tool.`,
@@ -263,8 +266,10 @@ export async function presentAssistantMessage(cline: Task) {
 
 				// Once a tool result has been collected, ignore all other tool
 				// uses since we should only ever present one tool result per
-				// message.
-				cline.didAlreadyUseTool = true
+				// message. Exception: update_todo_list can be used alongside other tools.
+				if (!isUpdateTodoList) {
+					cline.didAlreadyUseTool = true
+				}
 			}
 
 			const askApproval = async (