diff --git a/packages/types/src/__tests__/experiment.test.ts b/packages/types/src/__tests__/experiment.test.ts new file mode 100644 index 0000000000..0ef4ed2e02 --- /dev/null +++ b/packages/types/src/__tests__/experiment.test.ts @@ -0,0 +1,19 @@ +import { experimentIds, experimentIdsSchema, experimentsSchema } from "../experiment.js" + +describe("dynamicThinkingEffort experiment", () => { + it("is part of the experiment id enum", () => { + expect(experimentIds).toContain("dynamicThinkingEffort") + expect(experimentIdsSchema.safeParse("dynamicThinkingEffort").success).toBe(true) + }) + + it("parses enabled and disabled states", () => { + expect(experimentsSchema.parse({ dynamicThinkingEffort: true })).toEqual({ dynamicThinkingEffort: true }) + expect(experimentsSchema.parse({ dynamicThinkingEffort: false })).toEqual({ dynamicThinkingEffort: false }) + expect(experimentsSchema.parse({})).toEqual({}) + }) + + it("rejects non-boolean values", () => { + expect(experimentsSchema.safeParse({ dynamicThinkingEffort: "yes" }).success).toBe(false) + expect(experimentIdsSchema.safeParse("dynamic-thinking-effort").success).toBe(false) + }) +}) diff --git a/packages/types/src/experiment.ts b/packages/types/src/experiment.ts index 5d511859b1..f4b3a1c0a8 100644 --- a/packages/types/src/experiment.ts +++ b/packages/types/src/experiment.ts @@ -12,6 +12,7 @@ export const experimentIds = [ "runSlashCommand", "customTools", "parallelToolExecution", + "dynamicThinkingEffort", ] as const export const experimentIdsSchema = z.enum(experimentIds) @@ -28,6 +29,7 @@ export const experimentsSchema = z.object({ runSlashCommand: z.boolean().optional(), customTools: z.boolean().optional(), parallelToolExecution: z.boolean().optional(), + dynamicThinkingEffort: z.boolean().optional(), }) export type Experiments = z.infer diff --git a/scripts/stryker-diff.mjs b/scripts/stryker-diff.mjs index c0e8a6cd1a..e46afb22ab 100644 --- a/scripts/stryker-diff.mjs +++ b/scripts/stryker-diff.mjs @@ -51,8 +51,14 @@ export const PACKAGE_CONFIGS = [ root: "src", sourceRoot: "src/", vitestConfig: "vitest.config.ts", - vitestRelated: false, - discoverRelatedTests: true, + // Use plugin-side related-test discovery: with an explicit STRYKER_TEST_FILES + // list, stryker-js 10.0.0 plans static mutants with runtime activation + // (stryker-mutator/stryker-js#6144, #6209), which the vitest plugin can only + // apply after top-level code has run, producing false Survived results. + // Revert to vitestRelated: false / discoverRelatedTests: true once the + // upstream planner fix lands. + vitestRelated: true, + discoverRelatedTests: false, excludedPaths: ["src/esbuild.mjs", "src/eslint.config.mjs", "src/utils/vitest-verbosity.ts"], }, ] diff --git a/scripts/stryker-diff.test.mjs b/scripts/stryker-diff.test.mjs index 0f39dc507f..b86f5f3e96 100644 --- a/scripts/stryker-diff.test.mjs +++ b/scripts/stryker-diff.test.mjs @@ -129,8 +129,11 @@ describe("buildManifest", () => { assert.equal(webview.runRoot, ".") assert.equal(webview.discoverRelatedTests, true) assert.equal(webview.vitestRelated, false) - assert.equal(extension.discoverRelatedTests, true) - assert.equal(extension.vitestRelated, false) + // The extension entry uses plugin-side related discovery while stryker-js 10.0.0 + // plans static mutants with runtime activation when given explicit test files + // (stryker-mutator/stryker-js#6144, #6209); see the entry comment in stryker-diff.mjs. + assert.equal(extension.discoverRelatedTests, false) + assert.equal(extension.vitestRelated, true) }) it("returns no packages for tests, barrels, unsupported packages, and type-only changes", () => { diff --git a/src/api/index.ts b/src/api/index.ts index 98c3c5dc7b..ba1f4d7e0a 100644 --- a/src/api/index.ts +++ b/src/api/index.ts @@ -7,6 +7,7 @@ import { retiredProviderIdentifiers, type ProviderSettings, type ModelInfo, + type ReasoningEffortExtended, } from "@roo-code/types" import { getRouterRemovalMessage } from "../core/config/routerRemoval" @@ -115,6 +116,14 @@ export interface ApiHandlerCreateMessageMetadata { * when the user clicks stop, preventing wasted API tokens/compute on the provider side. */ abortSignal?: AbortSignal + /** + * Per-request thinking effort override (DTE series 2/5). + * When defined, takes precedence over the settings-derived `reasoningEffort` + * wherever the effective effort is resolved (see `resolveEffectiveReasoningEffort`). + * Task-scoped and transient: it applies to this request only (the next request + * after being set — no mid-stream effect) and is never persisted to settings. + */ + reasoningEffort?: ReasoningEffortExtended } export interface ApiHandler { diff --git a/src/api/transform/__tests__/dte-effective-reasoning-effort.spec.ts b/src/api/transform/__tests__/dte-effective-reasoning-effort.spec.ts new file mode 100644 index 0000000000..f126cae97c --- /dev/null +++ b/src/api/transform/__tests__/dte-effective-reasoning-effort.spec.ts @@ -0,0 +1,58 @@ +// npx vitest run src/api/transform/__tests__/dte-effective-reasoning-effort.spec.ts + +import { ADAPTIVE_OUTPUT_CONFIG_EFFORTS, resolveEffectiveReasoningEffort } from "../reasoning" + +describe("DTE series 2/5 — resolveEffectiveReasoningEffort", () => { + const settingsEffort = "high" + const modelDefault = "medium" + + it("returns the per-request override when present (strongest precedence)", () => { + expect( + resolveEffectiveReasoningEffort({ + override: "xhigh", + settingsReasoningEffort: settingsEffort, + modelDefaultEffort: modelDefault, + }), + ).toBe("xhigh") + }) + + it("lets the override win even when it is out-of-range for the adaptive envelope", () => { + // "minimal" is a valid override value but outside the adaptive envelope set; + // resolution still returns it — envelope gating is the caller's concern. + expect( + resolveEffectiveReasoningEffort({ + override: "minimal", + settingsReasoningEffort: settingsEffort, + modelDefaultEffort: modelDefault, + }), + ).toBe("minimal") + }) + + it("falls back to the settings value when no override is present", () => { + expect( + resolveEffectiveReasoningEffort({ settingsReasoningEffort: "low", modelDefaultEffort: modelDefault }), + ).toBe("low") + }) + + it("preserves the settings 'disable' sentinel when no override is present", () => { + expect( + resolveEffectiveReasoningEffort({ settingsReasoningEffort: "disable", modelDefaultEffort: modelDefault }), + ).toBe("disable") + }) + + it("an explicit override wins over a settings 'disable' sentinel", () => { + expect(resolveEffectiveReasoningEffort({ override: "low", settingsReasoningEffort: "disable" })).toBe("low") + }) + + it("falls back to the model default when neither override nor settings is set", () => { + expect(resolveEffectiveReasoningEffort({ modelDefaultEffort: "low" })).toBe("low") + }) + + it("returns undefined when nothing is set", () => { + expect(resolveEffectiveReasoningEffort({})).toBeUndefined() + }) + + it("exposes exactly the in-range adaptive envelope efforts", () => { + expect([...ADAPTIVE_OUTPUT_CONFIG_EFFORTS]).toEqual(["low", "medium", "high", "xhigh", "max"]) + }) +}) diff --git a/src/api/transform/reasoning.ts b/src/api/transform/reasoning.ts index c51111125a..14bdaba889 100644 --- a/src/api/transform/reasoning.ts +++ b/src/api/transform/reasoning.ts @@ -22,6 +22,51 @@ export type AnthropicProviderReasoningParams = AnthropicReasoningParams | { type export type OpenAiReasoningParams = { reasoning_effort: OpenAI.Chat.ChatCompletionCreateParams["reasoning_effort"] } +/** + * DTE series 2/5 — effort levels accepted by the Claude 4.7+ adaptive-thinking + * `output_config.effort` envelope. Efforts outside this set (e.g. "none", + * "minimal", "disable") omit the envelope so the API applies its own default. + */ +export const ADAPTIVE_OUTPUT_CONFIG_EFFORTS: readonly ReasoningEffortExtended[] = [ + "low", + "medium", + "high", + "xhigh", + "max", +] + +/** + * DTE series 2/5 — resolves the effective thinking effort for a single request. + * + * Resolution order (strongest first): + * 1. `override` — the per-request task-local effort + * (`ApiHandlerCreateMessageMetadata.reasoningEffort`), + * 2. `settingsReasoningEffort` — the settings-derived value, + * 3. `modelDefaultEffort` — the model's default effort. + * + * This is the single shared resolution point for the per-request override: + * providers that resolve the effective effort through it inherit the override + * without duplicating precedence logic. The override is transient (next request + * only) and never persisted to settings. + */ +export const resolveEffectiveReasoningEffort = ({ + override, + settingsReasoningEffort, + modelDefaultEffort, +}: { + override?: ReasoningEffortExtended + settingsReasoningEffort?: ReasoningEffortExtended | "disable" + modelDefaultEffort?: ReasoningEffortExtended +}): ReasoningEffortExtended | "disable" | undefined => { + if (override !== undefined) { + return override + } + if (settingsReasoningEffort !== undefined) { + return settingsReasoningEffort + } + return modelDefaultEffort +} + // Valid Gemini thinking levels for effort-based reasoning const GEMINI_THINKING_LEVELS = ["minimal", "low", "medium", "high"] as const diff --git a/src/core/webview/__tests__/ClineProvider.spec.ts b/src/core/webview/__tests__/ClineProvider.spec.ts index 1a6a82a5b0..5efd9ed769 100644 --- a/src/core/webview/__tests__/ClineProvider.spec.ts +++ b/src/core/webview/__tests__/ClineProvider.spec.ts @@ -1566,6 +1566,38 @@ describe("ClineProvider", () => { expect(state.destructiveCommandGuardEnabled).toBe(false) }) + test("getStateToPostToWebview returns the saved dynamicThinkingEffort experiment when enabled", async () => { + await provider.resolveWebviewView(mockWebviewView) + await provider.contextProxy.setValue("experiments", { + ...experimentDefault, + dynamicThinkingEffort: true, + }) + + const state = await provider.getStateToPostToWebview() + + expect(state.experiments).toEqual({ ...experimentDefault, dynamicThinkingEffort: true }) + }) + + test("getStateToPostToWebview returns the saved dynamicThinkingEffort experiment when explicitly disabled", async () => { + await provider.resolveWebviewView(mockWebviewView) + await provider.contextProxy.setValue("experiments", { + ...experimentDefault, + dynamicThinkingEffort: false, + }) + + const state = await provider.getStateToPostToWebview() + + expect(state.experiments).toEqual({ ...experimentDefault, dynamicThinkingEffort: false }) + }) + + test("getStateToPostToWebview defaults dynamicThinkingEffort to false when experiments are unset", async () => { + await provider.resolveWebviewView(mockWebviewView) + + const state = await provider.getStateToPostToWebview() + + expect(state.experiments).toEqual({ ...experimentDefault, dynamicThinkingEffort: false }) + }) + test("language is set to VSCode language", async () => { // Mock VSCode language as Spanish ;(vscode.env as any).language = "pt-BR" diff --git a/src/core/webview/__tests__/webviewMessageHandler.spec.ts b/src/core/webview/__tests__/webviewMessageHandler.spec.ts index 4c2a301965..bbbb6852e7 100644 --- a/src/core/webview/__tests__/webviewMessageHandler.spec.ts +++ b/src/core/webview/__tests__/webviewMessageHandler.spec.ts @@ -73,6 +73,7 @@ import type { ModelRecord } from "@roo-code/types" import { webviewMessageHandler } from "../webviewMessageHandler" import type { ClineProvider } from "../ClineProvider" +import { experimentDefault } from "../../../shared/experiments" import { flushModels, getModels } from "../../../api/providers/fetchers/modelCache" import { getLMStudioModels } from "../../../api/providers/fetchers/lmstudio" import { getCommands } from "../../../services/command/commands" @@ -2273,3 +2274,57 @@ describe("webviewMessageHandler - telemetrySetting", () => { expect(TelemetryService.instance.updateTelemetryState).not.toHaveBeenCalled() }) }) + +describe("webviewMessageHandler - experiments", () => { + beforeEach(() => { + vi.clearAllMocks() + }) + + it("persists dynamicThinkingEffort true merged with the saved experiments", async () => { + vi.mocked(mockClineProvider.contextProxy.getValue).mockImplementation((key: string) => + key === "experiments" ? { preventFocusDisruption: true } : undefined, + ) + + await webviewMessageHandler(mockClineProvider, { + type: "updateSettings", + updatedSettings: { experiments: { dynamicThinkingEffort: true } }, + }) + + expect(mockClineProvider.contextProxy.setValue).toHaveBeenCalledWith("experiments", { + preventFocusDisruption: true, + dynamicThinkingEffort: true, + }) + expect(mockClineProvider.postStateToWebview).toHaveBeenCalledTimes(1) + }) + + it("persists dynamicThinkingEffort false over a previously enabled value", async () => { + vi.mocked(mockClineProvider.contextProxy.getValue).mockImplementation((key: string) => + key === "experiments" ? { dynamicThinkingEffort: true } : undefined, + ) + + await webviewMessageHandler(mockClineProvider, { + type: "updateSettings", + updatedSettings: { experiments: { dynamicThinkingEffort: false } }, + }) + + expect(mockClineProvider.contextProxy.setValue).toHaveBeenCalledWith("experiments", { + dynamicThinkingEffort: false, + }) + expect(mockClineProvider.postStateToWebview).toHaveBeenCalledTimes(1) + }) + + it("persists the experiment defaults when no experiments are saved yet", async () => { + vi.mocked(mockClineProvider.contextProxy.getValue).mockReturnValue(undefined) + + await webviewMessageHandler(mockClineProvider, { + type: "updateSettings", + updatedSettings: { experiments: {} }, + }) + + expect(mockClineProvider.contextProxy.setValue).toHaveBeenCalledWith("experiments", { + ...experimentDefault, + dynamicThinkingEffort: false, + }) + expect(mockClineProvider.postStateToWebview).toHaveBeenCalledTimes(1) + }) +}) diff --git a/src/shared/__tests__/experiments.spec.ts b/src/shared/__tests__/experiments.spec.ts index f2261a5c09..b6e8993df4 100644 --- a/src/shared/__tests__/experiments.spec.ts +++ b/src/shared/__tests__/experiments.spec.ts @@ -2,7 +2,7 @@ import type { ExperimentId } from "@roo-code/types" -import { EXPERIMENT_IDS, experimentConfigsMap, experiments as Experiments } from "../experiments" +import { EXPERIMENT_IDS, experimentConfigsMap, experimentDefault, experiments as Experiments } from "../experiments" describe("experiments", () => { describe("PREVENT_FOCUS_DISRUPTION", () => { @@ -22,6 +22,7 @@ describe("experiments", () => { runSlashCommand: false, customTools: false, parallelToolExecution: false, + dynamicThinkingEffort: false, } expect(Experiments.isEnabled(experiments, EXPERIMENT_IDS.PREVENT_FOCUS_DISRUPTION)).toBe(false) }) @@ -33,6 +34,7 @@ describe("experiments", () => { runSlashCommand: false, customTools: false, parallelToolExecution: false, + dynamicThinkingEffort: false, } expect(Experiments.isEnabled(experiments, EXPERIMENT_IDS.PREVENT_FOCUS_DISRUPTION)).toBe(true) }) @@ -44,6 +46,7 @@ describe("experiments", () => { runSlashCommand: false, customTools: false, parallelToolExecution: false, + dynamicThinkingEffort: false, } expect(Experiments.isEnabled(experiments, EXPERIMENT_IDS.PREVENT_FOCUS_DISRUPTION)).toBe(false) }) @@ -66,4 +69,28 @@ describe("experiments", () => { expect(Experiments.isEnabled({ parallelToolExecution: true }, "parallelToolExecution")).toBe(true) }) }) + + describe("DYNAMIC_THINKING_EFFORT", () => { + it("is configured correctly", () => { + expect(EXPERIMENT_IDS.DYNAMIC_THINKING_EFFORT).toBe("dynamicThinkingEffort") + expect(experimentConfigsMap.DYNAMIC_THINKING_EFFORT).toMatchObject({ + enabled: false, + }) + // Visible in the Settings panel (showInSettings defaults to true). + expect(experimentConfigsMap.DYNAMIC_THINKING_EFFORT.showInSettings).toBeUndefined() + }) + + it("is disabled by default", () => { + expect(experimentDefault.dynamicThinkingEffort).toBe(false) + expect(Experiments.isEnabled({}, "dynamicThinkingEffort")).toBe(false) + }) + + it("returns true when enabled", () => { + expect(Experiments.isEnabled({ dynamicThinkingEffort: true }, "dynamicThinkingEffort")).toBe(true) + }) + + it("returns false when explicitly disabled", () => { + expect(Experiments.isEnabled({ dynamicThinkingEffort: false }, "dynamicThinkingEffort")).toBe(false) + }) + }) }) diff --git a/src/shared/experiments.ts b/src/shared/experiments.ts index ae538b9138..c0d461a454 100644 --- a/src/shared/experiments.ts +++ b/src/shared/experiments.ts @@ -6,6 +6,7 @@ export const EXPERIMENT_IDS = { RUN_SLASH_COMMAND: "runSlashCommand", CUSTOM_TOOLS: "customTools", PARALLEL_TOOL_EXECUTION: "parallelToolExecution", + DYNAMIC_THINKING_EFFORT: "dynamicThinkingEffort", } as const satisfies Record type _AssertExperimentIds = AssertEqual>> @@ -25,6 +26,7 @@ export const experimentConfigsMap: Record = { CUSTOM_TOOLS: { enabled: false }, // TODO: add i18n keys (settings:experimental.PARALLEL_TOOL_EXECUTION.name/.description) in the same PR that sets showInSettings: true PARALLEL_TOOL_EXECUTION: { enabled: false, showInSettings: false }, + DYNAMIC_THINKING_EFFORT: { enabled: false }, } export const experimentDefault = Object.fromEntries( diff --git a/webview-ui/playwright/gallery/stories.tsx b/webview-ui/playwright/gallery/stories.tsx index 04ca36b2c6..1cce748510 100644 --- a/webview-ui/playwright/gallery/stories.tsx +++ b/webview-ui/playwright/gallery/stories.tsx @@ -185,6 +185,11 @@ export const stories: Record = { const { UISettingsStory } = await import("@/components/settings/__tests__/UISettings.visual.fixture") return }, + "experimental-settings": async () => { + const { ExperimentalSettingsStory } = + await import("@/components/settings/__tests__/ExperimentalSettings.visual.fixture") + return + }, "ui-settings-long-locale": async () => { const [{ UISettingsStory }, { default: i18next }] = await Promise.all([ import("@/components/settings/__tests__/UISettings.visual.fixture"), diff --git a/webview-ui/src/components/settings/__tests__/ExperimentalSettings.spec.tsx b/webview-ui/src/components/settings/__tests__/ExperimentalSettings.spec.tsx index b31f87dc7e..3feddad1d4 100644 --- a/webview-ui/src/components/settings/__tests__/ExperimentalSettings.spec.tsx +++ b/webview-ui/src/components/settings/__tests__/ExperimentalSettings.spec.tsx @@ -1,4 +1,4 @@ -import { render, screen } from "@testing-library/react" +import { fireEvent, render, screen } from "@testing-library/react" import { experimentDefault } from "@roo/experiments" @@ -32,4 +32,78 @@ describe("ExperimentalSettings", () => { expect(screen.getByText("settings:experimental.CUSTOM_TOOLS.name")).toBeInTheDocument() expect(screen.queryByText("settings:experimental.PARALLEL_TOOL_EXECUTION.name")).not.toBeInTheDocument() }) + + it("renders the dynamic thinking effort toggle", () => { + render() + + expect(screen.getByText("settings:experimental.DYNAMIC_THINKING_EFFORT.name")).toBeInTheDocument() + }) + + it("leaves the dynamic thinking effort toggle unchecked when the value is false or omitted", () => { + const getCheckbox = () => { + const label = screen.getByText("settings:experimental.DYNAMIC_THINKING_EFFORT.name").closest("label") + return label?.querySelector("input[type='checkbox']") + } + + // Explicit false + let result = render( + , + ) + expect(getCheckbox()).not.toBeNull() + expect(getCheckbox()).not.toBeChecked() + result.unmount() + + // Omitted (absent from the persisted config) + const omitted: Record = { ...experimentDefault } + delete omitted.dynamicThinkingEffort + result = render() + expect(getCheckbox()).not.toBeNull() + expect(getCheckbox()).not.toBeChecked() + result.unmount() + }) + + it("binds the dynamic thinking effort toggle to setExperimentEnabled", () => { + const setExperimentEnabled = vi.fn() + render( + , + ) + + const label = screen.getByText("settings:experimental.DYNAMIC_THINKING_EFFORT.name").closest("label") + const checkbox = label?.querySelector("input[type='checkbox']") + expect(checkbox).not.toBeNull() + expect(checkbox).toBeChecked() + + fireEvent.click(checkbox!) + + expect(setExperimentEnabled).toHaveBeenCalledTimes(1) + expect(setExperimentEnabled).toHaveBeenCalledWith("dynamicThinkingEffort", false) + }) + + it("toggles the dynamic thinking effort on when clicked from the unchecked state", () => { + const setExperimentEnabled = vi.fn() + render( + , + ) + + const label = screen.getByText("settings:experimental.DYNAMIC_THINKING_EFFORT.name").closest("label") + const checkbox = label?.querySelector("input[type='checkbox']") + expect(checkbox).not.toBeNull() + expect(checkbox).not.toBeChecked() + + fireEvent.click(checkbox!) + + expect(setExperimentEnabled).toHaveBeenCalledTimes(1) + expect(setExperimentEnabled).toHaveBeenCalledWith("dynamicThinkingEffort", true) + }) }) diff --git a/webview-ui/src/components/settings/__tests__/ExperimentalSettings.visual.fixture.tsx b/webview-ui/src/components/settings/__tests__/ExperimentalSettings.visual.fixture.tsx new file mode 100644 index 0000000000..f28038aa6d --- /dev/null +++ b/webview-ui/src/components/settings/__tests__/ExperimentalSettings.visual.fixture.tsx @@ -0,0 +1,28 @@ +import { useState } from "react" + +import type { Experiments } from "@roo-code/types" + +import { EXPERIMENT_IDS, experimentDefault } from "@roo/experiments" + +import { ExperimentalSettings } from "../ExperimentalSettings" +import { AppProviders } from "../../../../playwright/AppProviders" + +export function ExperimentalSettingsStory() { + const [experiments, setExperiments] = useState({ + ...experimentDefault, + [EXPERIMENT_IDS.DYNAMIC_THINKING_EFFORT]: true, + }) + + return ( + +
+ setExperiments((current) => ({ ...current, [id]: enabled }))} + /> +
+ + ) +} diff --git a/webview-ui/src/components/settings/__tests__/ExperimentalSettings.visual.tsx b/webview-ui/src/components/settings/__tests__/ExperimentalSettings.visual.tsx new file mode 100644 index 0000000000..8a12ebf3ae --- /dev/null +++ b/webview-ui/src/components/settings/__tests__/ExperimentalSettings.visual.tsx @@ -0,0 +1,12 @@ +import { expect, test } from "../../../../playwright/coverage-fixture" +import { mountedStory } from "../../../../playwright/mounted-story" +import { applyVisualTheme, visualThemes } from "../../../../playwright/themes" + +for (const theme of visualThemes) { + test(`renders the experimental settings section in the VS Code ${theme.name} theme`, async ({ mount, page }) => { + const component = mountedStory(await mount("experimental-settings")) + await applyVisualTheme(page, theme) + const story = component.getByTestId("experimental-settings-story") + await expect(story).toHaveScreenshot(`experimental-settings-${theme.name}.png`) + }) +} diff --git a/webview-ui/src/components/settings/__tests__/SettingsView.spec.tsx b/webview-ui/src/components/settings/__tests__/SettingsView.spec.tsx index a3aa131902..656cf2e659 100644 --- a/webview-ui/src/components/settings/__tests__/SettingsView.spec.tsx +++ b/webview-ui/src/components/settings/__tests__/SettingsView.spec.tsx @@ -5,6 +5,7 @@ import { act } from "@testing-library/react" import { vscode } from "@/utils/vscode" import { DEFAULT_CHECKPOINT_TIMEOUT_SECONDS } from "@roo-code/types" +import { experimentDefault } from "@roo/experiments" import SettingsView from "../SettingsView" @@ -804,3 +805,50 @@ describe("SettingsView - Duplicate Commands", () => { expect(onDone).toHaveBeenCalledTimes(1) }) }) + +describe("SettingsView - Experimental Settings", () => { + beforeEach(() => { + vi.clearAllMocks() + }) + + it("includes the dynamic thinking effort toggle in the updateSettings payload on save", () => { + const { activateTab, getSettingsContent } = renderSettingsView({ experiments: experimentDefault }) + activateTab("experimental") + const content = getSettingsContent() + const label = within(content).getByText("settings:experimental.DYNAMIC_THINKING_EFFORT.name").closest("label")! + const checkbox = within(label).getByRole("checkbox") + expect(checkbox).not.toBeChecked() + fireEvent.click(checkbox) + fireEvent.click(screen.getByTestId("save-button")) + expect(vscode.postMessage).toHaveBeenCalledWith( + expect.objectContaining({ + type: "updateSettings", + updatedSettings: expect.objectContaining({ + experiments: expect.objectContaining({ dynamicThinkingEffort: true }), + }), + }), + ) + }) + + it("persists the unset representation of dynamicThinkingEffort as false on save", () => { + const experiments: Record = { ...experimentDefault } + delete experiments.dynamicThinkingEffort + const { activateTab, getSettingsContent } = renderSettingsView({ experiments }) + activateTab("experimental") + const content = getSettingsContent() + const label = within(content).getByText("settings:experimental.DYNAMIC_THINKING_EFFORT.name").closest("label")! + const checkbox = within(label).getByRole("checkbox") + expect(checkbox).not.toBeChecked() + fireEvent.click(checkbox) + fireEvent.click(checkbox) + fireEvent.click(screen.getByTestId("save-button")) + expect(vscode.postMessage).toHaveBeenCalledWith( + expect.objectContaining({ + type: "updateSettings", + updatedSettings: expect.objectContaining({ + experiments: expect.objectContaining({ dynamicThinkingEffort: false }), + }), + }), + ) + }) +}) diff --git a/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-dark.png b/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-dark.png new file mode 100644 index 0000000000..f7b15dc14b Binary files /dev/null and b/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-dark.png differ diff --git a/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-high-contrast-light.png b/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-high-contrast-light.png new file mode 100644 index 0000000000..4f1b090fdf Binary files /dev/null and b/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-high-contrast-light.png differ diff --git a/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-high-contrast.png b/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-high-contrast.png new file mode 100644 index 0000000000..348dab9629 Binary files /dev/null and b/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-high-contrast.png differ diff --git a/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-light.png b/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-light.png new file mode 100644 index 0000000000..a148e2c5d9 Binary files /dev/null and b/webview-ui/src/components/settings/__tests__/__screenshots__/experimental-settings-light.png differ diff --git a/webview-ui/src/i18n/locales/ca/settings.json b/webview-ui/src/i18n/locales/ca/settings.json index 762a1290af..8d7b7ee515 100644 --- a/webview-ui/src/i18n/locales/ca/settings.json +++ b/webview-ui/src/i18n/locales/ca/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Eines actualitzades correctament", "refreshError": "Error en actualitzar les eines", "toolParameters": "Paràmetres" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Esforç de pensament dinàmic", + "description": "Activa l'eina set_thinking_effort perquè el model decideixi el seu esforç de pensament per pas. El control manual al xat sempre està disponible. (experimental)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/de/settings.json b/webview-ui/src/i18n/locales/de/settings.json index f56984bc73..cbb94b5cee 100644 --- a/webview-ui/src/i18n/locales/de/settings.json +++ b/webview-ui/src/i18n/locales/de/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Tools erfolgreich aktualisiert", "refreshError": "Fehler beim Aktualisieren der Tools", "toolParameters": "Parameter" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Dynamische Denkintensität", + "description": "Aktiviert das set_thinking_effort-Tool, damit das Modell die Denkintensität pro Schritt selbst bestimmt. Die manuelle Anpassung im Chat ist immer verfügbar. (experimentell)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/en/settings.json b/webview-ui/src/i18n/locales/en/settings.json index 77f8a4f86d..9839de35d3 100644 --- a/webview-ui/src/i18n/locales/en/settings.json +++ b/webview-ui/src/i18n/locales/en/settings.json @@ -1055,6 +1055,10 @@ "refreshSuccess": "Tools refreshed successfully", "refreshError": "Failed to refresh tools", "toolParameters": "Parameters" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Dynamic thinking effort", + "description": "Enables the set_thinking_effort tool so the model can decide its thinking effort per step. The manual in-chat control is always available. (experimental)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/es/settings.json b/webview-ui/src/i18n/locales/es/settings.json index e730483fa0..9556372c04 100644 --- a/webview-ui/src/i18n/locales/es/settings.json +++ b/webview-ui/src/i18n/locales/es/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Herramientas actualizadas correctamente", "refreshError": "Error al actualizar las herramientas", "toolParameters": "Parámetros" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Esfuerzo de pensamiento dinámico", + "description": "Habilita la herramienta set_thinking_effort para que el modelo decida su esfuerzo de pensamiento por paso. El control manual en el chat siempre está disponible. (experimental)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/fr/settings.json b/webview-ui/src/i18n/locales/fr/settings.json index 78770da21d..2914341df1 100644 --- a/webview-ui/src/i18n/locales/fr/settings.json +++ b/webview-ui/src/i18n/locales/fr/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Outils actualisés avec succès", "refreshError": "Échec de l'actualisation des outils", "toolParameters": "Paramètres" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Effort de réflexion dynamique", + "description": "Active l'outil set_thinking_effort afin que le modèle décide de son effort de réflexion à chaque étape. Le contrôle manuel dans le chat est toujours disponible. (expérimental)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/hi/settings.json b/webview-ui/src/i18n/locales/hi/settings.json index 413d3515bc..9fb8a84a72 100644 --- a/webview-ui/src/i18n/locales/hi/settings.json +++ b/webview-ui/src/i18n/locales/hi/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "टूल्स सफलतापूर्वक रिफ्रेश हुए", "refreshError": "टूल्स रिफ्रेश करने में विफल", "toolParameters": "पैरामीटर्स" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "डायनामिक चिंतन प्रयास", + "description": "set_thinking_effort टूल सक्षम करता है ताकि मॉडल हर चरण में अपना चिंतन प्रयास स्वयं तय कर सके। चैट में मैन्युअल नियंत्रण हमेशा उपलब्ध है। (प्रयोगात्मक)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/id/settings.json b/webview-ui/src/i18n/locales/id/settings.json index 9b9928da64..682ac9ad55 100644 --- a/webview-ui/src/i18n/locales/id/settings.json +++ b/webview-ui/src/i18n/locales/id/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Tool berhasil direfresh", "refreshError": "Gagal merefresh tool", "toolParameters": "Parameter" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Usaha berpikir dinamis", + "description": "Mengaktifkan alat set_thinking_effort agar model dapat memutuskan usaha berpikirnya per langkah. Kontrol manual di obrolan selalu tersedia. (eksperimental)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/it/settings.json b/webview-ui/src/i18n/locales/it/settings.json index 8a43ec3d35..9ae8f2a8bd 100644 --- a/webview-ui/src/i18n/locales/it/settings.json +++ b/webview-ui/src/i18n/locales/it/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Strumenti aggiornati con successo", "refreshError": "Impossibile aggiornare gli strumenti", "toolParameters": "Parametri" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Sforzo di pensiero dinamico", + "description": "Abilita lo strumento set_thinking_effort in modo che il modello decida lo sforzo di pensiero per ogni passaggio. Il controllo manuale nella chat è sempre disponibile. (sperimentale)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/ja/settings.json b/webview-ui/src/i18n/locales/ja/settings.json index b2cfbe977e..b824141a95 100644 --- a/webview-ui/src/i18n/locales/ja/settings.json +++ b/webview-ui/src/i18n/locales/ja/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "ツールが正常に更新されました", "refreshError": "ツールの更新に失敗しました", "toolParameters": "パラメーター" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "ダイナミック思考強度", + "description": "set_thinking_effort ツールを有効にし、ステップごとにモデルが思考強度を決定できるようにします。チャット内での手動調整は常に利用可能です。（実験的機能）" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/ko/settings.json b/webview-ui/src/i18n/locales/ko/settings.json index 1ff9addeb4..c714cf86bf 100644 --- a/webview-ui/src/i18n/locales/ko/settings.json +++ b/webview-ui/src/i18n/locales/ko/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "도구가 성공적으로 새로고침되었습니다", "refreshError": "도구 새로고침에 실패했습니다", "toolParameters": "매개변수" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "동적 사고 노력", + "description": "set_thinking_effort 도구를 활성화하여 단계별로 모델이 사고 노력을 결정하도록 합니다. 채팅에서 수동 조정은 항상 사용 가능합니다. (실험적 기능)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/nl/settings.json b/webview-ui/src/i18n/locales/nl/settings.json index 4361d091a1..dfa0c29a7f 100644 --- a/webview-ui/src/i18n/locales/nl/settings.json +++ b/webview-ui/src/i18n/locales/nl/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Tools succesvol vernieuwd", "refreshError": "Fout bij vernieuwen van tools", "toolParameters": "Parameters" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Dynamisch denkwerk", + "description": "Schakelt het set_thinking_effort-gereedschap in zodat het model per stap het denkwerk zelf kan bepalen. De handmatige aanpassing in het gesprek is altijd beschikbaar. (experimenteel)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/pl/settings.json b/webview-ui/src/i18n/locales/pl/settings.json index 277bcaa470..84e06b68a6 100644 --- a/webview-ui/src/i18n/locales/pl/settings.json +++ b/webview-ui/src/i18n/locales/pl/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Narzędzia odświeżone pomyślnie", "refreshError": "Nie udało się odświeżyć narzędzi", "toolParameters": "Parametry" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Dynamiczny wysiłek myślowy", + "description": "Włącza narzędzie set_thinking_effort, aby model samodzielnie decydował o wysiłku myślowym na każdym kroku. Ręczna regulacja w czacie jest zawsze dostępna. (eksperymentalne)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/pt-BR/settings.json b/webview-ui/src/i18n/locales/pt-BR/settings.json index 8ce67bcd48..79e16a6298 100644 --- a/webview-ui/src/i18n/locales/pt-BR/settings.json +++ b/webview-ui/src/i18n/locales/pt-BR/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Ferramentas atualizadas com sucesso", "refreshError": "Falha ao atualizar ferramentas", "toolParameters": "Parâmetros" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Esforço de pensamento dinâmico", + "description": "Ativa a ferramenta set_thinking_effort para que o modelo decida seu esforço de pensamento em cada etapa. O controle manual na conversa está sempre disponível. (experimental)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/ru/settings.json b/webview-ui/src/i18n/locales/ru/settings.json index 0ac516c190..e1131173ff 100644 --- a/webview-ui/src/i18n/locales/ru/settings.json +++ b/webview-ui/src/i18n/locales/ru/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Инструменты успешно обновлены", "refreshError": "Не удалось обновить инструменты", "toolParameters": "Параметры" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Динамическое усилие размышления", + "description": "Включает инструмент set_thinking_effort, чтобы модель самостоятельно определяла усилие размышления на каждом шаге. Ручное управление в чате всегда доступно. (экспериментальная функция)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/tr/settings.json b/webview-ui/src/i18n/locales/tr/settings.json index 5d3a5cb89a..992ef00800 100644 --- a/webview-ui/src/i18n/locales/tr/settings.json +++ b/webview-ui/src/i18n/locales/tr/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Araçlar başarıyla yenilendi", "refreshError": "Araçlar yenilenemedi", "toolParameters": "Parametreler" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Dinamik düşünme çabası", + "description": "set_thinking_effort aracını etkinleştirir; böylece model her adımda kendi düşünme çabasını belirler. Sohbetteki manuel kontrol her zaman kullanılabilir. (deneysel)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/vi/settings.json b/webview-ui/src/i18n/locales/vi/settings.json index adb1be64e3..2f697a1609 100644 --- a/webview-ui/src/i18n/locales/vi/settings.json +++ b/webview-ui/src/i18n/locales/vi/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "Làm mới công cụ thành công", "refreshError": "Không thể làm mới công cụ", "toolParameters": "Thông số" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "Nỗ lực suy nghĩ động", + "description": "Bật công cụ set_thinking_effort để mô hình tự quyết định nỗ lực suy nghĩ của từng bước. Điều chỉnh thủ công trong trò chuyện luôn khả dụng. (thực nghiệm)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/zh-CN/settings.json b/webview-ui/src/i18n/locales/zh-CN/settings.json index 916f629efe..092ebfb8c3 100644 --- a/webview-ui/src/i18n/locales/zh-CN/settings.json +++ b/webview-ui/src/i18n/locales/zh-CN/settings.json @@ -975,6 +975,10 @@ "refreshSuccess": "工具刷新成功", "refreshError": "工具刷新失败", "toolParameters": "参数" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "动态思考强度", + "description": "启用 set_thinking_effort 工具，让模型按步骤自行决定思考强度。对话中的手动调整随时可用。 (实验性功能)" } }, "promptCaching": { diff --git a/webview-ui/src/i18n/locales/zh-TW/settings.json b/webview-ui/src/i18n/locales/zh-TW/settings.json index d1f9258dcf..a057fab671 100644 --- a/webview-ui/src/i18n/locales/zh-TW/settings.json +++ b/webview-ui/src/i18n/locales/zh-TW/settings.json @@ -1002,6 +1002,10 @@ "refreshSuccess": "工具重新整理成功", "refreshError": "工具重新整理失敗", "toolParameters": "參數" + }, + "DYNAMIC_THINKING_EFFORT": { + "name": "動態思考強度", + "description": "啟用 set_thinking_effort 工具，讓模型自行決定每一步的思考強度。對話中的手動調整隨時可用。（實驗性）" } }, "promptCaching": {