@machina.ai/cell-cli-core 1.11.0-rc1 → 1.13.0-rc1

package/dist/src/core/client.test.js

@@ -4,7 +4,7 @@
  * SPDX-License-Identifier: Apache-2.0
  */
 import { describe, it, expect, vi, beforeEach, afterEach, } from 'vitest';
-import { findCompressSplitPoint, isThinkingDefault, isThinkingSupported, GeminiClient, } from './client.js';
+import { isThinkingDefault, isThinkingSupported, GeminiClient, } from './client.js';
 import { AuthType, } from './contentGenerator.js';
 import {} from './geminiChat.js';
 import { CompressionStatus, GeminiEventType, Turn, } from './turn.js';
@@ -14,8 +14,9 @@ import { FileDiscoveryService } from '../services/fileDiscoveryService.js';
 import { setSimulate429 } from '../utils/testUtils.js';
 import { tokenLimit } from './tokenLimits.js';
 import { ideContextStore } from '../ide/ideContext.js';
-import { ClearcutLogger } from '../telemetry/clearcut-logger/clearcut-logger.js';
 import { uiTelemetryService } from '../telemetry/uiTelemetry.js';
+import { ChatCompressionService } from '../services/chatCompressionService.js';
+vi.mock('../services/chatCompressionService.js');
 // Mock fs module to prevent actual file system operations during tests
 const mockFileSystem = new Map();
 vi.mock('node:fs', () => {
@@ -95,70 +96,6 @@ async function fromAsync(promise) {
     }
     return results;
 }
-describe('findCompressSplitPoint', () => {
-    it('should throw an error for non-positive numbers', () => {
-        expect(() => findCompressSplitPoint([], 0)).toThrow('Fraction must be between 0 and 1');
-    });
-    it('should throw an error for a fraction greater than or equal to 1', () => {
-        expect(() => findCompressSplitPoint([], 1)).toThrow('Fraction must be between 0 and 1');
-    });
-    it('should handle an empty history', () => {
-        expect(findCompressSplitPoint([], 0.5)).toBe(0);
-    });
-    it('should handle a fraction in the middle', () => {
-        const history = [
-            { role: 'user', parts: [{ text: 'This is the first message.' }] }, // JSON length: 66 (19%)
-            { role: 'model', parts: [{ text: 'This is the second message.' }] }, // JSON length: 68 (40%)
-            { role: 'user', parts: [{ text: 'This is the third message.' }] }, // JSON length: 66 (60%)
-            { role: 'model', parts: [{ text: 'This is the fourth message.' }] }, // JSON length: 68 (80%)
-            { role: 'user', parts: [{ text: 'This is the fifth message.' }] }, // JSON length: 65 (100%)
-        ];
-        expect(findCompressSplitPoint(history, 0.5)).toBe(4);
-    });
-    it('should handle a fraction of last index', () => {
-        const history = [
-            { role: 'user', parts: [{ text: 'This is the first message.' }] }, // JSON length: 66 (19%)
-            { role: 'model', parts: [{ text: 'This is the second message.' }] }, // JSON length: 68 (40%)
-            { role: 'user', parts: [{ text: 'This is the third message.' }] }, // JSON length: 66 (60%)
-            { role: 'model', parts: [{ text: 'This is the fourth message.' }] }, // JSON length: 68 (80%)
-            { role: 'user', parts: [{ text: 'This is the fifth message.' }] }, // JSON length: 65 (100%)
-        ];
-        expect(findCompressSplitPoint(history, 0.9)).toBe(4);
-    });
-    it('should handle a fraction of after last index', () => {
-        const history = [
-            { role: 'user', parts: [{ text: 'This is the first message.' }] }, // JSON length: 66 (24%%)
-            { role: 'model', parts: [{ text: 'This is the second message.' }] }, // JSON length: 68 (50%)
-            { role: 'user', parts: [{ text: 'This is the third message.' }] }, // JSON length: 66 (74%)
-            { role: 'model', parts: [{ text: 'This is the fourth message.' }] }, // JSON length: 68 (100%)
-        ];
-        expect(findCompressSplitPoint(history, 0.8)).toBe(4);
-    });
-    it('should return earlier splitpoint if no valid ones are after threshhold', () => {
-        const history = [
-            { role: 'user', parts: [{ text: 'This is the first message.' }] },
-            { role: 'model', parts: [{ text: 'This is the second message.' }] },
-            { role: 'user', parts: [{ text: 'This is the third message.' }] },
-            { role: 'model', parts: [{ functionCall: {} }] },
-        ];
-        // Can't return 4 because the previous item has a function call.
-        expect(findCompressSplitPoint(history, 0.99)).toBe(2);
-    });
-    it('should handle a history with only one item', () => {
-        const historyWithEmptyParts = [
-            { role: 'user', parts: [{ text: 'Message 1' }] },
-        ];
-        expect(findCompressSplitPoint(historyWithEmptyParts, 0.5)).toBe(0);
-    });
-    it('should handle history with weird parts', () => {
-        const historyWithEmptyParts = [
-            { role: 'user', parts: [{ text: 'Message 1' }] },
-            { role: 'model', parts: [{ fileData: { fileUri: 'derp' } }] },
-            { role: 'user', parts: [{ text: 'Message 2' }] },
-        ];
-        expect(findCompressSplitPoint(historyWithEmptyParts, 0.5)).toBe(2);
-    });
-});
 describe('isThinkingSupported', () => {
     it('should return true for gemini-2.5', () => {
         expect(isThinkingSupported('gemini-2.5')).toBe(true);
@@ -194,6 +131,14 @@ describe('Gemini Client (client.ts)', () => {
     beforeEach(async () => {
         vi.resetAllMocks();
         vi.mocked(uiTelemetryService.setLastPromptTokenCount).mockClear();
+        vi.mocked(ChatCompressionService.prototype.compress).mockResolvedValue({
+            newHistory: null,
+            info: {
+                originalTokenCount: 0,
+                newTokenCount: 0,
+                compressionStatus: CompressionStatus.NOOP,
+            },
+        });
         mockGenerateContentFn = vi.fn().mockResolvedValue({
             candidates: [{ content: { parts: [{ text: '{"key": "value"}' }] } }],
         });
@@ -269,6 +214,7 @@ describe('Gemini Client (client.ts)', () => {
         client = new GeminiClient(mockConfig);
         await client.initialize();
         vi.mocked(mockConfig.getGeminiClient).mockReturnValue(client);
+        vi.mocked(uiTelemetryService.setLastPromptTokenCount).mockClear();
     });
     afterEach(() => {
         vi.restoreAllMocks();
@@ -319,87 +265,64 @@ describe('Gemini Client (client.ts)', () => {
                 getHistory: mockGetHistory,
                 addHistory: vi.fn(),
                 setHistory: vi.fn(),
+                getLastPromptTokenCount: vi.fn(),
             };
         });
         function setup({ chatHistory = [
             { role: 'user', parts: [{ text: 'Long conversation' }] },
             { role: 'model', parts: [{ text: 'Long response' }] },
-        ], originalTokenCount = 1000, summaryText = 'This is a summary.', } = {}) {
+        ], originalTokenCount = 1000, newTokenCount = 500, compressionStatus = CompressionStatus.COMPRESSED, } = {}) {
             const mockOriginalChat = {
                 getHistory: vi.fn((_curated) => chatHistory),
                 setHistory: vi.fn(),
+                getLastPromptTokenCount: vi.fn().mockReturnValue(originalTokenCount),
             };
             client['chat'] = mockOriginalChat;
             vi.mocked(uiTelemetryService.getLastPromptTokenCount).mockReturnValue(originalTokenCount);
-            mockGenerateContentFn.mockResolvedValue({
-                candidates: [
-                    {
-                        content: {
-                            role: 'model',
-                            parts: [{ text: summaryText }],
-                        },
-                    },
-                ],
-            });
-            // Calculate what the new history will be
-            const splitPoint = findCompressSplitPoint(chatHistory, 0.7); // 1 - 0.3
-            const historyToKeep = chatHistory.slice(splitPoint);
-            // This is the history that the new chat will have.
-            // It includes the default startChat history + the extra history from tryCompressChat
-            const newCompressedHistory = [
-                // Mocked envParts + canned response from startChat
-                {
-                    role: 'user',
-                    parts: [{ text: 'Mocked env context' }],
-                },
-                {
-                    role: 'model',
-                    parts: [{ text: 'Got it. Thanks for the context!' }],
-                },
-                // extraHistory from tryCompressChat
-                {
-                    role: 'user',
-                    parts: [{ text: summaryText }],
-                },
-                {
-                    role: 'model',
-                    parts: [{ text: 'Got it. Thanks for the additional context!' }],
-                },
-                ...historyToKeep,
+            const newHistory = [
+                { role: 'user', parts: [{ text: 'Summary' }] },
+                { role: 'model', parts: [{ text: 'Got it' }] },
             ];
+            vi.mocked(ChatCompressionService.prototype.compress).mockResolvedValue({
+                newHistory: compressionStatus === CompressionStatus.COMPRESSED
+                    ? newHistory
+                    : null,
+                info: {
+                    originalTokenCount,
+                    newTokenCount,
+                    compressionStatus,
+                },
+            });
             const mockNewChat = {
-                getHistory: vi.fn().mockReturnValue(newCompressedHistory),
+                getHistory: vi.fn().mockReturnValue(newHistory),
                 setHistory: vi.fn(),
+                getLastPromptTokenCount: vi.fn().mockReturnValue(newTokenCount),
             };
             client['startChat'] = vi
                 .fn()
                 .mockResolvedValue(mockNewChat);
-            const totalChars = newCompressedHistory.reduce((total, content) => total + JSON.stringify(content).length, 0);
-            const estimatedNewTokenCount = Math.floor(totalChars / 4);
             return {
                 client,
                 mockOriginalChat,
                 mockNewChat,
-                estimatedNewTokenCount,
+                estimatedNewTokenCount: newTokenCount,
             };
         }
         describe('when compression inflates the token count', () => {
             it('allows compression to be forced/manual after a failure', async () => {
-                // Call 1 (Fails): Setup with a long summary to inflate tokens
-                const longSummary = 'long summary '.repeat(100);
-                const { client, estimatedNewTokenCount: inflatedTokenCount } = setup({
+                // Call 1 (Fails): Setup with inflated tokens
+                setup({
                     originalTokenCount: 100,
-                    summaryText: longSummary,
+                    newTokenCount: 200,
+                    compressionStatus: CompressionStatus.COMPRESSION_FAILED_INFLATED_TOKEN_COUNT,
                 });
-                expect(inflatedTokenCount).toBeGreaterThan(100); // Ensure setup is correct
                 await client.tryCompressChat('prompt-id-4', false); // Fails
-                // Call 2 (Forced): Re-setup with a short summary
-                const shortSummary = 'short';
+                // Call 2 (Forced): Re-setup with compressed tokens
                 const { estimatedNewTokenCount: compressedTokenCount } = setup({
                     originalTokenCount: 100,
-                    summaryText: shortSummary,
+                    newTokenCount: 50,
+                    compressionStatus: CompressionStatus.COMPRESSED,
                 });
-                expect(compressedTokenCount).toBeLessThanOrEqual(100); // Ensure setup is correct
                 const result = await client.tryCompressChat('prompt-id-4', true); // Forced
                 expect(result).toEqual({
                     compressionStatus: CompressionStatus.COMPRESSED,
@@ -408,12 +331,11 @@ describe('Gemini Client (client.ts)', () => {
                 });
             });
             it('yields the result even if the compression inflated the tokens', async () => {
-                const longSummary = 'long summary '.repeat(100);
                 const { client, estimatedNewTokenCount } = setup({
                     originalTokenCount: 100,
-                    summaryText: longSummary,
+                    newTokenCount: 200,
+                    compressionStatus: CompressionStatus.COMPRESSION_FAILED_INFLATED_TOKEN_COUNT,
                 });
-                expect(estimatedNewTokenCount).toBeGreaterThan(100); // Ensure setup is correct
                 const result = await client.tryCompressChat('prompt-id-4', false);
                 expect(result).toEqual({
                     compressionStatus: CompressionStatus.COMPRESSION_FAILED_INFLATED_TOKEN_COUNT,
@@ -424,47 +346,52 @@ describe('Gemini Client (client.ts)', () => {
                 expect(uiTelemetryService.setLastPromptTokenCount).not.toHaveBeenCalled();
             });
             it('does not manipulate the source chat', async () => {
-                const longSummary = 'long summary '.repeat(100);
-                const { client, mockOriginalChat, estimatedNewTokenCount } = setup({
+                const { client, mockOriginalChat } = setup({
                     originalTokenCount: 100,
-                    summaryText: longSummary,
+                    newTokenCount: 200,
+                    compressionStatus: CompressionStatus.COMPRESSION_FAILED_INFLATED_TOKEN_COUNT,
                 });
-                expect(estimatedNewTokenCount).toBeGreaterThan(100); // Ensure setup is correct
                 await client.tryCompressChat('prompt-id-4', false);
                 // On failure, the chat should NOT be replaced
                 expect(client['chat']).toBe(mockOriginalChat);
             });
-            it('will not attempt to compress context after a failure', async () => {
-                const longSummary = 'long summary '.repeat(100);
-                const { client, estimatedNewTokenCount } = setup({
+            it.skip('will not attempt to compress context after a failure', async () => {
+                const { client } = setup({
                     originalTokenCount: 100,
-                    summaryText: longSummary,
+                    newTokenCount: 200,
+                    compressionStatus: CompressionStatus.COMPRESSION_FAILED_INFLATED_TOKEN_COUNT,
                 });
-                expect(estimatedNewTokenCount).toBeGreaterThan(100); // Ensure setup is correct
                 await client.tryCompressChat('prompt-id-4', false); // This fails and sets hasFailedCompressionAttempt = true
+                // Mock the next call to return NOOP
+                vi.mocked(ChatCompressionService.prototype.compress).mockResolvedValueOnce({
+                    newHistory: null,
+                    info: {
+                        originalTokenCount: 0,
+                        newTokenCount: 0,
+                        compressionStatus: CompressionStatus.NOOP,
+                    },
+                });
                 // This call should now be a NOOP
                 const result = await client.tryCompressChat('prompt-id-5', false);
-                // generateContent (for summary) should only have been called once
-                expect(mockGenerateContentFn).toHaveBeenCalledTimes(1);
-                expect(result).toEqual({
-                    compressionStatus: CompressionStatus.NOOP,
-                    newTokenCount: 0,
-                    originalTokenCount: 0,
-                });
+                expect(result.compressionStatus).toBe(CompressionStatus.NOOP);
+                expect(ChatCompressionService.prototype.compress).toHaveBeenCalledTimes(2);
+                expect(ChatCompressionService.prototype.compress).toHaveBeenLastCalledWith(expect.anything(), 'prompt-id-5', false, expect.anything(), expect.anything(), true);
             });
         });
         it('should not trigger summarization if token count is below threshold', async () => {
             const MOCKED_TOKEN_LIMIT = 1000;
-            vi.mocked(tokenLimit).mockReturnValue(MOCKED_TOKEN_LIMIT);
-            mockGetHistory.mockReturnValue([
-                { role: 'user', parts: [{ text: '...history...' }] },
-            ]);
             const originalTokenCount = MOCKED_TOKEN_LIMIT * 0.699;
-            vi.mocked(uiTelemetryService.getLastPromptTokenCount).mockReturnValue(originalTokenCount);
+            vi.mocked(ChatCompressionService.prototype.compress).mockResolvedValue({
+                newHistory: null,
+                info: {
+                    originalTokenCount,
+                    newTokenCount: originalTokenCount,
+                    compressionStatus: CompressionStatus.NOOP,
+                },
+            });
             const initialChat = client.getChat();
             const result = await client.tryCompressChat('prompt-id-2', false);
             const newChat = client.getChat();
-            expect(tokenLimit).toHaveBeenCalled();
             expect(result).toEqual({
                 compressionStatus: CompressionStatus.NOOP,
                 newTokenCount: originalTokenCount,
@@ -476,6 +403,8 @@ describe('Gemini Client (client.ts)', () => {
             const { client } = setup({
                 chatHistory: [{ role: 'user', parts: [{ text: 'hi' }] }],
                 originalTokenCount: 50,
+                newTokenCount: 50,
+                compressionStatus: CompressionStatus.NOOP,
             });
             const result = await client.tryCompressChat('prompt-id-noop', false);
             expect(result).toEqual({
@@ -483,270 +412,6 @@ describe('Gemini Client (client.ts)', () => {
                 originalTokenCount: 50,
                 newTokenCount: 50,
             });
-            expect(mockGenerateContentFn).not.toHaveBeenCalled();
-        });
-        it('logs a telemetry event when compressing', async () => {
-            vi.spyOn(ClearcutLogger.prototype, 'logChatCompressionEvent');
-            const MOCKED_TOKEN_LIMIT = 1000;
-            const MOCKED_CONTEXT_PERCENTAGE_THRESHOLD = 0.5;
-            vi.spyOn(client['config'], 'getChatCompression').mockReturnValue({
-                contextPercentageThreshold: MOCKED_CONTEXT_PERCENTAGE_THRESHOLD,
-            });
-            const history = [
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-            ];
-            mockGetHistory.mockReturnValue(history);
-            const originalTokenCount = MOCKED_TOKEN_LIMIT * MOCKED_CONTEXT_PERCENTAGE_THRESHOLD;
-            vi.mocked(uiTelemetryService.getLastPromptTokenCount).mockReturnValue(originalTokenCount);
-            // We need to control the estimated new token count.
-            // We mock startChat to return a chat with a known history.
-            const summaryText = 'This is a summary.';
-            const splitPoint = findCompressSplitPoint(history, 0.7);
-            const historyToKeep = history.slice(splitPoint);
-            const newCompressedHistory = [
-                { role: 'user', parts: [{ text: 'Mocked env context' }] },
-                { role: 'model', parts: [{ text: 'Got it. Thanks for the context!' }] },
-                { role: 'user', parts: [{ text: summaryText }] },
-                {
-                    role: 'model',
-                    parts: [{ text: 'Got it. Thanks for the additional context!' }],
-                },
-                ...historyToKeep,
-            ];
-            const mockNewChat = {
-                getHistory: vi.fn().mockReturnValue(newCompressedHistory),
-            };
-            client['startChat'] = vi
-                .fn()
-                .mockResolvedValue(mockNewChat);
-            const totalChars = newCompressedHistory.reduce((total, content) => total + JSON.stringify(content).length, 0);
-            const newTokenCount = Math.floor(totalChars / 4);
-            // Mock the summary response from the chat
-            mockGenerateContentFn.mockResolvedValue({
-                candidates: [
-                    {
-                        content: {
-                            role: 'model',
-                            parts: [{ text: summaryText }],
-                        },
-                    },
-                ],
-            });
-            await client.tryCompressChat('prompt-id-3', false);
-            expect(ClearcutLogger.prototype.logChatCompressionEvent).toHaveBeenCalledWith(expect.objectContaining({
-                tokens_before: originalTokenCount,
-                tokens_after: newTokenCount,
-            }));
-            expect(uiTelemetryService.setLastPromptTokenCount).toHaveBeenCalledWith(newTokenCount);
-            expect(uiTelemetryService.setLastPromptTokenCount).toHaveBeenCalledTimes(1);
-        });
-        it('should trigger summarization if token count is at threshold with contextPercentageThreshold setting', async () => {
-            const MOCKED_TOKEN_LIMIT = 1000;
-            const MOCKED_CONTEXT_PERCENTAGE_THRESHOLD = 0.5;
-            vi.mocked(tokenLimit).mockReturnValue(MOCKED_TOKEN_LIMIT);
-            vi.spyOn(client['config'], 'getChatCompression').mockReturnValue({
-                contextPercentageThreshold: MOCKED_CONTEXT_PERCENTAGE_THRESHOLD,
-            });
-            const history = [
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-            ];
-            mockGetHistory.mockReturnValue(history);
-            const originalTokenCount = MOCKED_TOKEN_LIMIT * MOCKED_CONTEXT_PERCENTAGE_THRESHOLD;
-            vi.mocked(uiTelemetryService.getLastPromptTokenCount).mockReturnValue(originalTokenCount);
-            // Mock summary and new chat
-            const summaryText = 'This is a summary.';
-            const splitPoint = findCompressSplitPoint(history, 0.7);
-            const historyToKeep = history.slice(splitPoint);
-            const newCompressedHistory = [
-                { role: 'user', parts: [{ text: 'Mocked env context' }] },
-                { role: 'model', parts: [{ text: 'Got it. Thanks for the context!' }] },
-                { role: 'user', parts: [{ text: summaryText }] },
-                {
-                    role: 'model',
-                    parts: [{ text: 'Got it. Thanks for the additional context!' }],
-                },
-                ...historyToKeep,
-            ];
-            const mockNewChat = {
-                getHistory: vi.fn().mockReturnValue(newCompressedHistory),
-            };
-            client['startChat'] = vi
-                .fn()
-                .mockResolvedValue(mockNewChat);
-            const totalChars = newCompressedHistory.reduce((total, content) => total + JSON.stringify(content).length, 0);
-            const newTokenCount = Math.floor(totalChars / 4);
-            // Mock the summary response from the chat
-            mockGenerateContentFn.mockResolvedValue({
-                candidates: [
-                    {
-                        content: {
-                            role: 'model',
-                            parts: [{ text: summaryText }],
-                        },
-                    },
-                ],
-            });
-            const initialChat = client.getChat();
-            const result = await client.tryCompressChat('prompt-id-3', false);
-            const newChat = client.getChat();
-            expect(tokenLimit).toHaveBeenCalled();
-            expect(mockGenerateContentFn).toHaveBeenCalled();
-            // Assert that summarization happened and returned the correct stats
-            expect(result).toEqual({
-                compressionStatus: CompressionStatus.COMPRESSED,
-                originalTokenCount,
-                newTokenCount,
-            });
-            // Assert that the chat was reset
-            expect(newChat).not.toBe(initialChat);
-        });
-        it('should not compress across a function call response', async () => {
-            const MOCKED_TOKEN_LIMIT = 1000;
-            vi.mocked(tokenLimit).mockReturnValue(MOCKED_TOKEN_LIMIT);
-            const history = [
-                { role: 'user', parts: [{ text: '...history 1...' }] },
-                { role: 'model', parts: [{ text: '...history 2...' }] },
-                { role: 'user', parts: [{ text: '...history 3...' }] },
-                { role: 'model', parts: [{ text: '...history 4...' }] },
-                { role: 'user', parts: [{ text: '...history 5...' }] },
-                { role: 'model', parts: [{ text: '...history 6...' }] },
-                { role: 'user', parts: [{ text: '...history 7...' }] },
-                { role: 'model', parts: [{ text: '...history 8...' }] },
-                // Normally we would break here, but we have a function response.
-                {
-                    role: 'user',
-                    parts: [{ functionResponse: { name: '...history 8...' } }],
-                },
-                { role: 'model', parts: [{ text: '...history 10...' }] },
-                // Instead we will break here.
-                { role: 'user', parts: [{ text: '...history 10...' }] },
-            ];
-            mockGetHistory.mockReturnValue(history);
-            const originalTokenCount = 1000 * 0.7;
-            vi.mocked(uiTelemetryService.getLastPromptTokenCount).mockReturnValue(originalTokenCount);
-            // Mock summary and new chat
-            const summaryText = 'This is a summary.';
-            const splitPoint = findCompressSplitPoint(history, 0.7); // This should be 10
-            expect(splitPoint).toBe(10); // Verify split point logic
-            const historyToKeep = history.slice(splitPoint); // Should keep last user message
-            expect(historyToKeep).toEqual([
-                { role: 'user', parts: [{ text: '...history 10...' }] },
-            ]);
-            const newCompressedHistory = [
-                { role: 'user', parts: [{ text: 'Mocked env context' }] },
-                { role: 'model', parts: [{ text: 'Got it. Thanks for the context!' }] },
-                { role: 'user', parts: [{ text: summaryText }] },
-                {
-                    role: 'model',
-                    parts: [{ text: 'Got it. Thanks for the additional context!' }],
-                },
-                ...historyToKeep,
-            ];
-            const mockNewChat = {
-                getHistory: vi.fn().mockReturnValue(newCompressedHistory),
-            };
-            client['startChat'] = vi
-                .fn()
-                .mockResolvedValue(mockNewChat);
-            const totalChars = newCompressedHistory.reduce((total, content) => total + JSON.stringify(content).length, 0);
-            const newTokenCount = Math.floor(totalChars / 4);
-            // Mock the summary response from the chat
-            mockGenerateContentFn.mockResolvedValue({
-                candidates: [
-                    {
-                        content: {
-                            role: 'model',
-                            parts: [{ text: summaryText }],
-                        },
-                    },
-                ],
-            });
-            const initialChat = client.getChat();
-            const result = await client.tryCompressChat('prompt-id-3', false);
-            const newChat = client.getChat();
-            expect(tokenLimit).toHaveBeenCalled();
-            expect(mockGenerateContentFn).toHaveBeenCalled();
-            // Assert that summarization happened and returned the correct stats
-            expect(result).toEqual({
-                compressionStatus: CompressionStatus.COMPRESSED,
-                originalTokenCount,
-                newTokenCount,
-            });
-            // Assert that the chat was reset
-            expect(newChat).not.toBe(initialChat);
-            // 1. standard start context message (env)
-            // 2. standard canned model response
-            // 3. compressed summary message (user)
-            // 4. standard canned model response
-            // 5. The last user message (historyToKeep)
-            expect(newChat.getHistory().length).toEqual(5);
-        });
-        it('should always trigger summarization when force is true, regardless of token count', async () => {
-            const history = [
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-                { role: 'user', parts: [{ text: '...history...' }] },
-                { role: 'model', parts: [{ text: '...history...' }] },
-            ];
-            mockGetHistory.mockReturnValue(history);
-            const originalTokenCount = 100; // Well below threshold, but > estimated new count
-            vi.mocked(uiTelemetryService.getLastPromptTokenCount).mockReturnValue(originalTokenCount);
-            // Mock summary and new chat
-            const summaryText = 'This is a summary.';
-            const splitPoint = findCompressSplitPoint(history, 0.7);
-            const historyToKeep = history.slice(splitPoint);
-            const newCompressedHistory = [
-                { role: 'user', parts: [{ text: 'Mocked env context' }] },
-                { role: 'model', parts: [{ text: 'Got it. Thanks for the context!' }] },
-                { role: 'user', parts: [{ text: summaryText }] },
-                {
-                    role: 'model',
-                    parts: [{ text: 'Got it. Thanks for the additional context!' }],
-                },
-                ...historyToKeep,
-            ];
-            const mockNewChat = {
-                getHistory: vi.fn().mockReturnValue(newCompressedHistory),
-            };
-            client['startChat'] = vi
-                .fn()
-                .mockResolvedValue(mockNewChat);
-            const totalChars = newCompressedHistory.reduce((total, content) => total + JSON.stringify(content).length, 0);
-            const newTokenCount = Math.floor(totalChars / 4);
-            // Mock the summary response from the chat
-            mockGenerateContentFn.mockResolvedValue({
-                candidates: [
-                    {
-                        content: {
-                            role: 'model',
-                            parts: [{ text: summaryText }],
-                        },
-                    },
-                ],
-            });
-            const initialChat = client.getChat();
-            const result = await client.tryCompressChat('prompt-id-1', true); // force = true
-            const newChat = client.getChat();
-            expect(mockGenerateContentFn).toHaveBeenCalled();
-            expect(result).toEqual({
-                compressionStatus: CompressionStatus.COMPRESSED,
-                originalTokenCount,
-                newTokenCount,
-            });
-            // Assert that the chat was reset
-            expect(newChat).not.toBe(initialChat);
         });
     });
     describe('sendMessageStream', () => {
@@ -831,6 +496,7 @@ describe('Gemini Client (client.ts)', () => {
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             const initialRequest = [{ text: 'Hi' }];
@@ -878,6 +544,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             const initialRequest = [{ text: 'Hi' }];
@@ -922,6 +589,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             const initialRequest = [{ text: 'Hi' }];
@@ -982,6 +650,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             const initialRequest = [{ text: 'Hi' }];
@@ -1015,6 +684,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             // Act
@@ -1048,6 +718,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             // Use a signal that never gets aborted
@@ -1093,6 +764,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             // Act & Assert
@@ -1131,6 +803,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             // Use a signal that never gets aborted
@@ -1158,9 +831,8 @@ ${JSON.stringify({
                     }
                 }
             }
-            catch (error) {
+            catch (_) {
                 // If the test framework times out, that also demonstrates the infinite loop
-                console.error('Test timed out or errored:', error);
             }
             // Assert that the fix works - the loop should stop at MAX_TURNS
             const callCount = mockCheckNextSpeaker.mock.calls.length;
@@ -1175,7 +847,11 @@ ${JSON.stringify({
             vi.mocked(tokenLimit).mockReturnValue(MOCKED_TOKEN_LIMIT);
             // Set last prompt token count
             const lastPromptTokenCount = 900;
-            vi.mocked(uiTelemetryService.getLastPromptTokenCount).mockReturnValue(lastPromptTokenCount);
+            const mockChat = {
+                getLastPromptTokenCount: vi.fn().mockReturnValue(lastPromptTokenCount),
+                getHistory: vi.fn().mockReturnValue([]),
+            };
+            client['chat'] = mockChat;
             // Remaining = 100. Threshold (95%) = 95.
             // We need a request > 95 tokens.
             // A string of length 400 is roughly 100 tokens.
@@ -1218,7 +894,11 @@ ${JSON.stringify({
             client['currentSequenceModel'] = STICKY_MODEL;
             // Set token count
             const lastPromptTokenCount = 900;
-            vi.mocked(uiTelemetryService.getLastPromptTokenCount).mockReturnValue(lastPromptTokenCount);
+            const mockChat = {
+                getLastPromptTokenCount: vi.fn().mockReturnValue(lastPromptTokenCount),
+                getHistory: vi.fn().mockReturnValue([]),
+            };
+            client['chat'] = mockChat;
             // Remaining (sticky) = 100. Threshold (95%) = 95.
             // We need a request > 95 tokens.
             const longText = 'a'.repeat(400);
@@ -1257,6 +937,12 @@ ${JSON.stringify({
                 mockTurnRunFn.mockReturnValue((async function* () {
                     yield { type: 'content', value: 'Hello' };
                 })());
+                const mockChat = {
+                    addHistory: vi.fn(),
+                    getHistory: vi.fn().mockReturnValue([]),
+                    getLastPromptTokenCount: vi.fn(),
+                };
+                client['chat'] = mockChat;
             });
             it('should use the model router service to select a model on the first turn', async () => {
                 const stream = client.sendMessageStream([{ text: 'Hi' }], new AbortController().signal, 'prompt-1');
@@ -1345,6 +1031,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             const initialRequest = [{ text: 'Hi' }];
@@ -1375,6 +1062,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             const initialRequest = [{ text: 'Hi' }];
@@ -1398,6 +1086,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             const initialRequest = [{ text: 'Hi' }];
@@ -1435,6 +1124,7 @@ ${JSON.stringify({
                         .mockReturnValue([
                         { role: 'user', parts: [{ text: 'previous message' }] },
                     ]),
+                    getLastPromptTokenCount: vi.fn(),
                 };
                 client['chat'] = mockChat;
             });
@@ -1555,7 +1245,11 @@ ${JSON.stringify({
                 vi.mocked(ideContextStore.get).mockReturnValue({
                     workspaceState: {
                         openFiles: [
-                            { ...currentActiveFile, isActive: true, timestamp: Date.now() },
+                            {
+                                ...currentActiveFile,
+                                isActive: true,
+                                timestamp: Date.now(),
+                            },
                         ],
                     },
                 });
@@ -1643,6 +1337,7 @@ ${JSON.stringify({
                     addHistory: vi.fn(),
                     getHistory: vi.fn().mockReturnValue([]), // Default empty history
                     setHistory: vi.fn(),
+                    getLastPromptTokenCount: vi.fn(),
                 };
                 client['chat'] = mockChat;
                 vi.spyOn(client['config'], 'getIdeMode').mockReturnValue(true);
@@ -1906,6 +1601,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             // Act
@@ -1931,6 +1627,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             // Act
@@ -1958,6 +1655,7 @@ ${JSON.stringify({
             const mockChat = {
                 addHistory: vi.fn(),
                 getHistory: vi.fn().mockReturnValue([]),
+                getLastPromptTokenCount: vi.fn(),
             };
             client['chat'] = mockChat;
             // Act