chore(release): refresh v1.0.3-beta.3

Author: zerob13

CHANGELOG.md

@@ -1,14 +1,22 @@
 # Changelog
 
-## v1.0.3-beta.3 (2026-04-17)
+## v1.0.3-beta.3 (2026-04-18)
 - Added Anthropic temperature support in model capability controls
 - Added `none` and `xhigh` reasoning effort options for supported models
 - Improved sidebar session pin feedback and stabilized session group identity handling
 - Refined app update installation by cleaning up floating windows before relaunch
+- Added Astraflow (ModelVerse) provider support and removed the deprecated Laoshi provider
+- Added project-based workspace directories with drag-and-drop setup support
+- Enhanced the floating agent widget to support all agents with more stable session handling
+- Improved Anthropic reasoning routing and capped derived max token defaults for safer model setup
 - 为模型能力配置补充 Anthropic temperature 支持
 - 为受支持模型新增 `none` 与 `xhigh` reasoning effort 选项
 - 优化侧栏会话 pin 反馈，并稳定会话分组标识处理
 - 在应用更新安装前清理悬浮窗口，提升升级流程稳定性
+- 新增 Astraflow（ModelVerse）Provider 支持，并移除已废弃的 Laoshi Provider
+- 新增基于项目目录的工作区管理能力，并支持拖拽接入工作区
+- 增强悬浮 Agent 按钮，支持全部 Agent 并提升会话管理稳定性
+- 优化 Anthropic reasoning 路由，并限制推导出的默认 max tokens，降低模型配置风险
 
 ## v1.0.3-beta.2 (2026-04-15)
 - Expanded remote control into a unified multi-channel setup with Discord, QQ Bot, and WeChat iLink support

test/main/presenter/configPresenter/anthropicProviderMigration.test.ts

@@ -1,6 +1,4 @@
 import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest'
-import { eventBus } from '@/eventbus'
-import { CONFIG_EVENTS } from '../../../../src/main/events'
 
 vi.mock('@/eventbus', () => ({
   eventBus: {
@@ -153,72 +151,3 @@ describe('getAnthropicModelSelectionKeysToClear', () => {
     expect(keysToClear).toEqual([])
   })
 })
-
-describe('migrateLegacyDefaultVisionModelToBuiltinAgent', () => {
-  beforeEach(() => {
-    vi.clearAllMocks()
-  })
-
-  it('migrates a valid legacy vision model with trimmed ids', () => {
-    const store = {
-      get: vi.fn().mockReturnValue({ providerId: ' openai ', modelId: ' gpt-4o ' }),
-      delete: vi.fn()
-    }
-    const updateBuiltinDeepChatConfig = vi.fn()
-    const presenter = Object.assign(Object.create(ConfigPresenter.prototype), {
-      store,
-      getBuiltinDeepChatConfig: vi.fn().mockReturnValue({}),
-      updateBuiltinDeepChatConfig
-    })
-
-    expect(() =>
-      (
-        presenter as ConfigPresenter & {
-          migrateLegacyDefaultVisionModelToBuiltinAgent: () => void
-        }
-      ).migrateLegacyDefaultVisionModelToBuiltinAgent()
-    ).not.toThrow()
-
-    expect(updateBuiltinDeepChatConfig).toHaveBeenCalledWith({
-      visionModel: {
-        providerId: 'openai',
-        modelId: 'gpt-4o'
-      }
-    })
-    expect(store.delete).toHaveBeenCalledWith('defaultVisionModel')
-    expect(eventBus.sendToMain).toHaveBeenCalledWith(
-      CONFIG_EVENTS.SETTING_CHANGED,
-      'defaultVisionModel',
-      undefined
-    )
-  })
-
-  it('cleans up malformed legacy vision model without throwing', () => {
-    const store = {
-      get: vi.fn().mockReturnValue({ providerId: { bad: true }, modelId: 'gpt-4o' }),
-      delete: vi.fn()
-    }
-    const updateBuiltinDeepChatConfig = vi.fn()
-    const presenter = Object.assign(Object.create(ConfigPresenter.prototype), {
-      store,
-      getBuiltinDeepChatConfig: vi.fn().mockReturnValue({}),
-      updateBuiltinDeepChatConfig
-    })
-
-    expect(() =>
-      (
-        presenter as ConfigPresenter & {
-          migrateLegacyDefaultVisionModelToBuiltinAgent: () => void
-        }
-      ).migrateLegacyDefaultVisionModelToBuiltinAgent()
-    ).not.toThrow()
-
-    expect(updateBuiltinDeepChatConfig).not.toHaveBeenCalled()
-    expect(store.delete).toHaveBeenCalledWith('defaultVisionModel')
-    expect(eventBus.sendToMain).toHaveBeenCalledWith(
-      CONFIG_EVENTS.SETTING_CHANGED,
-      'defaultVisionModel',
-      undefined
-    )
-  })
-})

test/main/presenter/configPresenter/providerModelCapabilityMapping.test.ts

@@ -125,7 +125,7 @@ describe('ConfigPresenter provider model capability mapping', () => {
     expect(supportsReasoning).toHaveBeenCalledWith('anthropic', 'claude-opus-4-7')
   })
 
-  it('maps anthropic transport relays to anthropic capability semantics', async () => {
+  it('keeps anthropic transport relays on provider-local capability semantics', async () => {
     const { ConfigPresenter, modelCapabilities } = await loadConfigPresenter()
     const supportsReasoning = vi
       .spyOn(modelCapabilities, 'supportsReasoning')
@@ -161,10 +161,10 @@ describe('ConfigPresenter provider model capability mapping', () => {
     expect(models).toEqual([
       expect.objectContaining({
         id: 'claude-opus-4-7',
-        reasoning: true
+        reasoning: false
       })
     ])
-    expect(supportsReasoning).toHaveBeenCalledWith('anthropic', 'claude-opus-4-7')
+    expect(supportsReasoning).not.toHaveBeenCalled()
   })
 
   it('maps zenmux anthropic routes to anthropic capability semantics', async () => {

test/renderer/components/ModelConfigDialog.test.ts

@@ -439,7 +439,7 @@ describe('ModelConfigDialog reasoning portraits', () => {
     expect(wrapper.text()).toContain('settings.model.modelConfig.reasoningVisibility.label')
   })
 
-  it('treats anthropic transport relays as editable anthropic toggles with conditional subsettings', async () => {
+  it('keeps anthropic transport relays on provider-local reasoning controls', async () => {
     const { wrapper } = await setup({
       providerId: 'my-anthropic-proxy',
       modelId: 'claude-opus-4-7',
@@ -466,17 +466,12 @@ describe('ModelConfigDialog reasoning portraits', () => {
       }
     })
 
-    expect((wrapper.vm as any).reasoningToggleMode).toBe('toggle')
-    expect((wrapper.vm as any).showReasoningEffort).toBe(false)
-    expect((wrapper.vm as any).showReasoningVisibility).toBe(false)
-
-    ;(wrapper.vm as any).config.reasoning = true
-    await nextTick()
-
+    expect((wrapper.vm as any).reasoningToggleMode).toBe('indicator')
+    expect((wrapper.vm as any).reasoningToggleDisabled).toBe(true)
+    expect((wrapper.vm as any).reasoningToggleValue).toBe(true)
     expect((wrapper.vm as any).showReasoningEffort).toBe(true)
-    expect((wrapper.vm as any).showReasoningVisibility).toBe(true)
-    expect((wrapper.vm as any).config.reasoningVisibility).toBe('omitted')
-    expect(wrapper.text()).toContain('settings.model.modelConfig.reasoningVisibility.label')
+    expect((wrapper.vm as any).showReasoningVisibility).toBe(false)
+    expect(wrapper.text()).not.toContain('settings.model.modelConfig.reasoningVisibility.label')
   })
 
   it('hides effort and budget controls for level-based portraits', async () => {

test/renderer/components/WindowSideBarSessionItem.test.ts

@@ -103,7 +103,7 @@ describe('WindowSideBarSessionItem', () => {
     expect(wrapper.find('[aria-label="thread.actions.pin"]').exists()).toBe(true)
   }, 10000)
 
-  it('exposes hero transition and pin feedback state on the rendered item', async () => {
+  it('exposes hero transition class and pin feedback state on the rendered item', async () => {
     const wrapper = await mountComponent({
       isPinned: true,
       heroHidden: true,
@@ -114,9 +114,8 @@ describe('WindowSideBarSessionItem', () => {
 
     expect(item.attributes('data-pin-fx')).toBe('pinning')
     expect(item.attributes('data-session-id')).toBe('session-1')
-    expect(item.attributes('data-hero-hidden')).toBe('true')
+    expect(item.classes()).toContain('is-hero-hidden')
     expect(item.attributes('data-pin-state')).toBe('docked')
-    expect(item.attributes('data-pin-visual-state')).toBe('overlay')
   }, 10000)
 
   it('keeps the pin layout docked while unpinning feedback is active', async () => {
@@ -126,7 +125,6 @@ describe('WindowSideBarSessionItem', () => {
     })
 
     expect(wrapper.find('.session-item').attributes('data-pin-state')).toBe('docked')
-    expect(wrapper.find('.session-item').attributes('data-pin-visual-state')).toBe('docked')
   }, 10000)
 
   it('highlights matching title fragments when filtering the sidebar', async () => {