[pull] main from microsoft:main

.eslint-plugin-local/code-no-telemetry-common-property.ts

@@ -49,6 +49,7 @@ const commonTelemetryProperties = new Set([
 	'common.copilottrackingid',
 	'common.copilotsdkversion',
 	'common.copilotruntimeversion',
+	'copilotsku',
 	'common.isagentswindow',
 ]);
 

build/azure-pipelines/product-build-ado-ci.yml

@@ -297,7 +297,7 @@ extends:
 
                   if [ "${#trailer_values[@]}" = "0" ]; then
                     echo "##vso[task.logissue type=error]Commit $head_sha is missing the required 'Msrc-Case-Id' trailer."
-                    printf "Add a trailer to the commit message, for example:\n\n    Msrc-Case-Id: 12345\n\n"
+                    printf "Add a trailer to the commit message, for example:\n\n    Msrc-Case-Id: 12345\n\nIf there is no associated case ID, use N/A:\n\n    Msrc-Case-Id: N/A\n\n"
                     exit 1
                   fi
 

build/lib/policies/policyData.jsonc

@@ -38,6 +38,21 @@
         }
     ],
     "policies": [
+        {
+            "key": "agents.voice.enabled",
+            "name": "AgentsVoice",
+            "category": "InteractiveSession",
+            "minimumVersion": "1.137",
+            "localization": {
+                "description": {
+                    "key": "agents.voice.enabled",
+                    "value": "Enable the Voice Mode panel in the chat view for voice-driven coding conversations."
+                }
+            },
+            "type": "boolean",
+            "default": false,
+            "included": true
+        },
         {
             "key": "chat.agent.allowedNetworkDomains",
             "name": "ChatAgentAllowedNetworkDomains",

extensions/copilot/package.json

@@ -5242,16 +5242,6 @@
               "onExp"
             ]
           },
-          "github.copilot.chat.agentDebugLog.enabled": {
-            "type": "boolean",
-            "default": false,
-            "markdownDescription": "%github.copilot.config.chat.agentDebugLog.enabled%",
-            "deprecationMessage": "%github.copilot.config.chat.agentDebugLog.enabled.deprecated%",
-            "tags": [
-              "advanced",
-              "experimental"
-            ]
-          },
           "github.copilot.chat.agentDebugLog.fileLogging.enabled": {
             "type": "boolean",
             "default": false,

extensions/copilot/package.nls.json

@@ -390,8 +390,6 @@
 	"github.copilot.config.inlineChat.reasoningEffort": "Controls the reasoning effort level for inline chat requests. Lower values result in faster responses with fewer reasoning tokens. Supported values depend on the model.",
 	"github.copilot.config.inlineChat.enableThinking": "Controls whether thinking/reasoning is enabled for inline chat requests. When disabled, reasoning summaries are suppressed for faster responses.",
 	"github.copilot.config.debug.requestLogger.maxEntries": "Maximum number of entries to keep in the request logger for debugging purposes.",
-	"github.copilot.config.chat.agentDebugLog.enabled": "Deprecated: use `github.copilot.chat.agentDebugLog.fileLogging.enabled` instead.",
-	"github.copilot.config.chat.agentDebugLog.enabled.deprecated": "This setting has been merged into `github.copilot.chat.agentDebugLog.fileLogging.enabled`. Please use this setting instead.",
 	"github.copilot.config.chat.agentDebugLog.fileLogging.enabled": "Enable agent debug logging: write chat debug events (tool calls, LLM requests, token usage, errors) to JSONL files for the debug panel and troubleshoot skill. Requires window reload to take effect.",
 	"github.copilot.config.chat.agentDebugLog.fileLogging.flushIntervalMs": "How often (in milliseconds) buffered debug log entries are flushed to disk. Lower values provide more up-to-date logs at the cost of more frequent disk writes.",
 	"github.copilot.config.chat.agentDebugLog.fileLogging.maxRetainedSessionLogs": "Maximum number of chat debug session log directories to retain on disk. Each chat session produces one directory. Older session logs are automatically deleted when this limit is exceeded.",

extensions/copilot/src/platform/configuration/common/configurationService.ts

@@ -846,9 +846,7 @@ export namespace ConfigKey {
 		/** Simulate GitHub authentication failures for testing. Can't be TeamInternal because we lose these flags as part of testing. */
 		export const DebugGitHubAuthFailWith = defineSetting<'NotAuthorized' | 'RequestFailed' | 'ParseFailed' | 'HTTP401' | 'RateLimited' | 'GitHubLoginFailed' | null>('chat.debug.githubAuthFailWith', ConfigType.Simple, null);
 
-		// Agent debug logging settings — fileLogging.enabled is the canonical toggle
-		/** @deprecated Use ChatDebugFileLogging instead. Kept during experiment transition. */
-		export const AgentDebugLogEnabled = defineAndMigrateExpSetting<boolean>('agentDebugLog.enabled', 'chat.agentDebugLog.enabled', false);
+		// Agent debug logging settings
 		export const ChatDebugFileLogging = defineAndMigrateExpSetting<boolean>('chat.chatDebug.fileLogging.enabled', 'chat.agentDebugLog.fileLogging.enabled', false);
 		export const ChatDebugFileLoggingFlushInterval = defineAndMigrateSetting<number>('chat.chatDebug.fileLogging.flushIntervalMs', 'chat.agentDebugLog.fileLogging.flushIntervalMs', 4000);
 		export const ChatDebugFileLoggingMaxRetainedSessionLogs = defineSetting<number>('chat.agentDebugLog.fileLogging.maxRetainedSessionLogs', ConfigType.ExperimentBased, 50);

extensions/copilot/src/platform/endpoint/common/autoModeTiers.ts

@@ -40,10 +40,11 @@ const retiredAutoModeTiers: Readonly<Record<string, AutoModeTier>> = {
 
 /**
  * Maps a retired tier name to its current one, leaving anything else untouched.
- * Only raw inputs that bypass the picker schema need this, notably the override setting.
+ * Needed for raw inputs that predate the rename, such as the override setting or a persisted picker
+ * value restored before its model's schema has loaded.
  */
-export function normalizeAutoModeTier(value: string): string {
-	return retiredAutoModeTiers[value] ?? value;
+export function normalizeAutoModeTier(value: unknown): unknown {
+	return typeof value === 'string' ? retiredAutoModeTiers[value] ?? value : value;
 }
 
 /**

extensions/copilot/src/platform/endpoint/node/automodeService.ts

@@ -404,19 +404,17 @@ export class AutomodeService extends Disposable implements IAutomodeService {
 	private _resolveTier(chatRequest: IAutoModeRoutingRequest | undefined): AutoModeTier | undefined {
 		const override = this._configurationService.getConfig(ConfigKey.Advanced.AutoModeTierOverride);
 		if (override) {
-			// Normalized because the override is a raw string setting, so unlike a picker value it never
-			// passes through the schema filter that drops retired names.
 			const normalized = normalizeAutoModeTier(override);
 			// The override is internal, so unlike the picker it may select `fast`.
-			if ((autoModeTiers as readonly string[]).includes(normalized)) {
+			if (autoModeTiers.some(tier => tier === normalized)) {
 				return normalized as AutoModeTier;
 			}
 			this._logService.warn(`[AutomodeService] Ignoring auto tier override '${override}' — not one of [${autoModeTiers.join(', ')}].`);
 		}
 		if (!this.areAutoModeTiersSupported()) {
 			return undefined;
 		}
-		const configured = chatRequest?.modelConfiguration?.[AUTO_MODE_TIER_PROPERTY];
+		const configured = normalizeAutoModeTier(chatRequest?.modelConfiguration?.[AUTO_MODE_TIER_PROPERTY]);
 		if (isSelectableAutoModeTier(configured) && configured !== defaultAutoModeTier) {
 			return configured;
 		}

extensions/copilot/src/platform/endpoint/node/test/automodeService.spec.ts

@@ -870,6 +870,32 @@ describe('AutomodeService', () => {
 
 			expect(autoRequestBodies()).toEqual([{ prompt: 'panel turn', tier: 'efficiency' }]);
 		});
+
+		// A picker value stored before the rename can be restored unfiltered while its model's
+		// schema is still loading, so it must upgrade rather than silently fall back.
+		it('maps a retired tier name in a persisted picker value to its current one', async () => {
+			enableTiers();
+			const gpt4oEndpoint = createEndpoint('gpt-4o', 'OpenAI');
+			mockAuto(autoResponse('gpt-4o'));
+
+			automodeService = createService();
+			await automodeService.resolveAutoModeEndpoint({
+				location: ChatLocation.Panel,
+				prompt: 'panel turn',
+				sessionId: 'session-persisted-retired',
+				modelConfiguration: { tier: 'max' },
+			} as unknown as ChatRequest, [mockChatEndpoint, gpt4oEndpoint]);
+			// `balanced` upgrades to the current default, which reads as "never picked", so the inline
+			// pin applies instead of being treated as an explicit selection.
+			await automodeService.resolveAutoModeEndpoint({
+				location: ChatLocation.Editor,
+				prompt: 'inline turn',
+				sessionId: 'session-persisted-retired-default',
+				modelConfiguration: { tier: 'balanced' },
+			} as unknown as ChatRequest, [mockChatEndpoint, gpt4oEndpoint]);
+
+			expect(autoRequestBodies().map(b => b.tier)).toEqual(['intelligence', 'fast']);
+		});
 	});
 
 	describe('session cache', () => {

extensions/copilot/src/platform/networking/node/fetcherFallback.ts

@@ -20,6 +20,7 @@ const fetcherConfigKeys: Partial<Record<FetcherId, Config<boolean>>> = {
 const terminalResponseStatusCodes = new Set([429, 502, 503]);
 const allFetchersFailedTelemetryIntervalMs = 15 * 60 * 1000;
 const maxAggregatedErrorLength = 1024;
+const maxAggregatedErrorCount = 100;
 const maxAggregatedErrorsSerializedLength = 8192;
 const aggregatedErrorsOverflowKey = '<other>';
 const allFetchersFailedErrors = new Map<string, number>();
@@ -85,7 +86,12 @@ function recordErrors(target: Map<string, number>, errors: readonly string[]): v
 			target.set(truncatedError, Math.min(count + 1, Number.MAX_SAFE_INTEGER));
 			continue;
 		}
-		target.set(truncatedError, 1);
+		if (target.size < maxAggregatedErrorCount) {
+			target.set(truncatedError, 1);
+			continue;
+		}
+		const overflowCount = target.get(aggregatedErrorsOverflowKey) ?? 0;
+		target.set(aggregatedErrorsOverflowKey, Math.min(overflowCount + 1, Number.MAX_SAFE_INTEGER));
 	}
 }
 

extensions/copilot/src/platform/networking/test/node/fetcherFallback.spec.ts

@@ -296,6 +296,66 @@ suite('FetcherFallback Test Suite', function () {
 		}
 	});
 
+	test('caps terminal response error cardinality while preserving counts', async function () {
+		vi.useFakeTimers();
+		cleanupTime += 24 * 60 * 60 * 1000;
+		vi.setSystemTime(cleanupTime);
+		const primingTelemetryService = new SpyingTelemetryService();
+		const primingFetchers = createTestFetchers([
+			{ name: 'fetcher1', response: createFakeResponse(503, someJSON) },
+			{ name: 'fetcher2', response: createFakeResponse(200, someJSON) },
+		]);
+		await fetchWithFallbacks(primingFetchers.fetchers, 'https://example.com', { callSite: 'test', expectJSON: true, retryFallbacks: true }, knownBadFetchers, configurationService, logService, primingTelemetryService, experimentationService);
+		vi.advanceTimersByTime(16 * 60 * 1000);
+		const createSuccessfulFetchers = () => createTestFetchers([
+			{ name: 'fetcher1', response: createFakeResponse(200, someJSON) },
+			{ name: 'fetcher2', response: createFakeResponse(200, someJSON) },
+		]);
+		await fetchWithFallbacks(createSuccessfulFetchers().fetchers, 'https://example.com', { callSite: 'test', expectJSON: true, retryFallbacks: true }, knownBadFetchers, configurationService, logService, primingTelemetryService, experimentationService);
+		const spyingTelemetryService = new SpyingTelemetryService();
+		try {
+			for (let i = 0; i < 200; i++) {
+				const testFetchers = createTestFetchers([
+					{ name: 'fetcher1', response: createFakeResponse(503, someJSON, `unique status ${i}`) },
+					{ name: 'fetcher2', response: createFakeResponse(200, someJSON) },
+				]);
+				await fetchWithFallbacks(testFetchers.fetchers, 'https://example.com', { callSite: 'test', expectJSON: true, retryFallbacks: true }, knownBadFetchers, configurationService, logService, spyingTelemetryService, experimentationService);
+			}
+			const repeatedOverflowFetchers = createTestFetchers([
+				{ name: 'fetcher1', response: createFakeResponse(503, someJSON, 'unique status 199') },
+				{ name: 'fetcher2', response: createFakeResponse(200, someJSON) },
+			]);
+			await fetchWithFallbacks(repeatedOverflowFetchers.fetchers, 'https://example.com', { callSite: 'test', expectJSON: true, retryFallbacks: true }, knownBadFetchers, configurationService, logService, spyingTelemetryService, experimentationService);
+
+			vi.advanceTimersByTime(15 * 60 * 1000 + 1);
+			await fetchWithFallbacks(createSuccessfulFetchers().fetchers, 'https://example.com', { callSite: 'test', expectJSON: true, retryFallbacks: true }, knownBadFetchers, configurationService, logService, spyingTelemetryService, experimentationService);
+			const telemetryEvents = spyingTelemetryService.getEvents().telemetryServiceEvents;
+			const properties = telemetryEvents[0].properties;
+			if (!properties || !('errors' in properties) || typeof properties.errors !== 'string') {
+				assert.fail('Expected an errors telemetry property');
+			}
+			const errors: Record<string, number> = JSON.parse(properties.errors);
+
+			assert.deepStrictEqual({
+				eventCount: telemetryEvents.length,
+				eventName: telemetryEvents[0].eventName,
+				errorKeyCount: Object.keys(errors).length,
+				overflowCount: errors['<other>'],
+				repeatedOverflowErrorCount: errors['fetcher1: 503 unique status 199'],
+				allFailuresAccountedFor: Object.values(errors).reduce((total, count) => total + count, 0),
+			}, {
+				eventCount: 1,
+				eventName: 'fetcherTerminalResponse',
+				errorKeyCount: 101,
+				overflowCount: 101,
+				repeatedOverflowErrorCount: undefined,
+				allFailuresAccountedFor: 201,
+			});
+		} finally {
+			vi.useRealTimers();
+		}
+	});
+
 	test('no fetcher succeeds', async function () {
 		const fetcherSpec = [
 			{ name: 'fetcher1', response: createFakeResponse(407, someHTML) },
@@ -407,7 +467,7 @@ function createTestFetchers(fetcherSpecs: Array<{ name: string; response: Respon
 				}
 				return next;
 			},
-			fetchWithPagination: async <T>(baseUrl: string, options: PaginationOptions<T>): Promise<T[]> => {
+			fetchWithPagination: async <T>(_baseUrl: string, _options: PaginationOptions<T>): Promise<T[]> => {
 				throw new Error('Method not implemented.');
 			},
 			disconnectAll: async () => { },
@@ -422,10 +482,10 @@ function createTestFetchers(fetcherSpecs: Array<{ name: string; response: Respon
 	return { fetchers, calls };
 }
 
-function createFakeResponse(statusCode: number, content: string) {
+function createFakeResponse(statusCode: number, content: string, statusText = 'status text') {
 	return Response.fromText(
 		statusCode,
-		'status text',
+		statusText,
 		new FakeHeaders(),
 		content,
 		'test-stub'

extensions/html-language-features/package.json

@@ -224,13 +224,6 @@
           "default": true,
           "description": "%html.hover.references%"
         },
-        "html.mirrorCursorOnMatchingTag": {
-          "type": "boolean",
-          "scope": "resource",
-          "default": false,
-          "description": "%html.mirrorCursorOnMatchingTag%",
-          "deprecationMessage": "%html.mirrorCursorOnMatchingTagDeprecationMessage%"
-        },
         "html.trace.server": {
           "type": "string",
           "scope": "window",

extensions/html-language-features/package.nls.json

@@ -33,8 +33,6 @@
 	"html.completion.attributeDefaultValue.doublequotes": "Attribute value is set to \"\".",
 	"html.completion.attributeDefaultValue.singlequotes": "Attribute value is set to ''.",
 	"html.completion.attributeDefaultValue.empty": "Attribute value is not set.",
-	"html.mirrorCursorOnMatchingTag": "Enable/disable mirroring cursor on matching HTML tag.",
-	"html.mirrorCursorOnMatchingTagDeprecationMessage": "Deprecated in favor of `editor.linkedEditing`",
 	"html.hover.documentation": "Show tag and attribute documentation in hover.",
 	"html.hover.references": "Show references to MDN in hover."
 }

extensions/npm/README.md

@@ -15,7 +15,7 @@ For more information about auto detection of Tasks, see the [documentation](http
 
 ### Script Explorer
 
-The Npm Script Explorer shows the npm scripts found in your workspace. The explorer view is enabled by the setting `npm.enableScriptExplorer`. A script can be opened, run, or debug from the explorer.
+The Npm Script Explorer shows the npm scripts found in your workspace. A script can be opened, run, or debug from the explorer.
 
 ### Run Scripts from the Editor
 
@@ -37,7 +37,6 @@ The extension fetches data from <https://registry.npmjs.org> and <https://regist
 - `npm.packageManager` - The package manager used to install dependencies: `auto`, `npm`, `yarn`, `pnpm` or `bun`. The default is `auto`, which detects your package manager based on files in your workspace.
 - `npm.scriptRunner` - The script runner used to run the scripts: `auto`, `npm`, `yarn`, `pnpm`, `bun` or `node`. The default is `auto`, which detects your script runner based on files in your workspace.
 - `npm.exclude` - Glob patterns for folders that should be excluded from automatic script detection. The pattern is matched against the **absolute path** of the package.json. For example, to exclude all test folders use '&ast;&ast;/test/&ast;&ast;'.
-- `npm.enableScriptExplorer` - Enable an explorer view for npm scripts.
 - `npm.scriptExplorerAction` - The default click action: `open` or `run`, the default is `open`.
 - `npm.enableRunFromFolder` - Enable running npm scripts from the context menu of folders in Explorer, the default is `false`.
 - `npm.scriptCodeLens.enable` - Enable/disable the code lenses to run a script, the default is `false`.

extensions/npm/package.json

@@ -292,13 +292,6 @@
           "description": "%config.npm.exclude%",
           "scope": "resource"
         },
-        "npm.enableScriptExplorer": {
-          "type": "boolean",
-          "default": false,
-          "scope": "resource",
-          "deprecationMessage": "The NPM Script Explorer is now available in 'Views' menu in the Explorer in all folders.",
-          "markdownDescription": "%config.npm.enableScriptExplorer%"
-        },
         "npm.enableRunFromFolder": {
           "type": "boolean",
           "default": false,

extensions/npm/package.nls.json

@@ -20,7 +20,6 @@
 	"config.npm.scriptRunner.vp": "Use Vite+ (vp) as the script runner.",
 	"config.npm.scriptRunner.auto": "Auto-detect which script runner to use based on lock files and installed package managers.",
 	"config.npm.exclude": "Configure glob patterns for folders that should be excluded from automatic script detection.",
-	"config.npm.enableScriptExplorer": "Enable an explorer view for npm scripts when there is no top-level `package.json` file.",
 	"config.npm.scriptExplorerAction": "The default click action used in the NPM Scripts Explorer: `open` or `run`, the default is `open`.",
 	"config.npm.scriptExplorerExclude": "An array of regular expressions that indicate which scripts should be excluded from the NPM Scripts view.",
 	"config.npm.enableRunFromFolder": "Enable running npm scripts contained in a folder from the Explorer context menu.",