64 Commits

This Branch

This Branch

All Branches

• fix: mandatory login enforcement and robust electron detection
  Gemini AI · 2025-12-29 02:11:21 +04:00
  0ee6903a8e
• fix: ReferenceError join in user-store.ts
  Gemini AI · 2025-12-29 02:04:57 +04:00
  97620aaaef
• fix: ensure all API requests use user ID and fix missing data migration for existing users
  Gemini AI · 2025-12-29 01:58:15 +04:00
  57720e6c5b
• fix: resolve JSX closing tag error in App.tsx
  Gemini AI · 2025-12-29 01:32:32 +04:00
  39d1e03785
• fix: import logoutActiveUser in main.ts
  Gemini AI · 2025-12-29 01:27:39 +04:00
  97a9b0955b
• feat: implement mandatory login on startup and set roman password
  Gemini AI · 2025-12-29 01:25:43 +04:00
  dc21ade84e
• feat: complete per-user integration config isolation and UI initialization
  Gemini AI · 2025-12-29 01:13:31 +04:00
  8474be8559
• feat: add API Status Checker tab in Advanced Settings with connection status and navigation
  Gemini AI · 2025-12-29 00:50:27 +04:00
  721da6f2ee
• fix: add robust Z.AI config persistence with error handling and logging
  Gemini AI · 2025-12-29 00:13:55 +04:00
  229f86c229
• fix: ensure Z.AI config directory exists before writing
  Gemini AI · 2025-12-28 12:30:45 +04:00
  2ac7eb12ce
• chore: exclude Dev Chat logs from tracking
  Gemini AI · 2025-12-28 11:28:55 +04:00
  0d3de9efa7
• chore: remove sensitive session data and progress logs from tracking
  Gemini AI · 2025-12-28 11:27:09 +04:00
  0d8913c788
• chore: minor fixes in session-actions and session-api
  Gemini AI · 2025-12-28 11:19:39 +04:00
  41f38fe4bf
• fix: resolve Antigravity token definition error and add validation

  Changes:
  1. Exported getStoredAntigravityToken and isAntigravityTokenValid from session-api.ts
  2. Imported token helpers into session-actions.ts
  3. Added token validation and user notifications to streamAntigravityChat
  4. Fixed TypeScript implicit any error in fetchAntigravityProvider

  Gemini AI · 2025-12-28 04:01:58 +04:00
  c66bd68a13
• fix: correctly track model provider ID for Qwen and all providers

  Changes:
  1. Added cachedProviderId signal in MultiX v2 component
  2. Updated syncFromStore to also sync provider ID from task session
  3. Immediately update cached model and provider IDs on model change
  4. Pass correct providerId to LiteModelSelector component
  
  This fixes the issue where selecting Qwen models caused chat input to stop
  responding, because the provider ID was not being tracked correctly.

  Gemini AI · 2025-12-28 03:40:56 +04:00
  6bd329701b
• fix: prevent duplicate AI models in selector and fix TypeScript errors

  Changes:
  1. Enhanced removeDuplicateProviders() to filter out duplicate providers from SDK
     when the same provider exists in extras (qwen-oauth, zai, ollama-cloud, antigravity)
  2. Added logic to remove any Qwen-related SDK providers when qwen-oauth is authenticated
  3. Fixed missing setActiveParentSession import in instance-shell2.tsx
  
  These changes ensure:
  - No duplicate models appear in the model selector
  - Qwen OAuth models don't duplicate with any SDK Qwen providers
  - TypeScript compilation passes successfully

  Gemini AI · 2025-12-28 03:27:31 +04:00
  38cb8bcb0c
• chore: user modifications
  Gemini AI · 2025-12-28 01:03:54 +04:00
  babce0e0a9
• fix: robust installers for all platforms (Windows, Linux, macOS)

  Windows:
  - Multiple Node.js installation methods (winget, chocolatey, direct MSI)
  - Clear restart instructions when Node.js is newly installed
  - Fallback to user profile directory if current dir not writable
  - Comprehensive health checks and error reporting
  
  Linux:
  - Support for apt, dnf, yum, pacman, zypper, apk package managers
  - NodeSource repository for newer Node.js versions
  - nvm fallback installation method
  - Graceful error handling with detailed troubleshooting
  
  macOS:
  - Xcode Command Line Tools detection and installation
  - Homebrew auto-installation
  - Apple Silicon (arm64) support with correct PATH setup
  - Multiple Node.js fallbacks (brew, nvm, official pkg)
  
  All platforms:
  - Binary-Free Mode as default (no OpenCode binary required)
  - Beautiful terminal output with progress indicators
  - Detailed logging to install.log
  - Post-install health checks

  Gemini AI · 2025-12-28 00:43:08 +04:00
  74001c7c3e
• fix: complete session persistence overhaul (Codex 5.2)

  1. Implemented auto-selection of tasks in MultiXV2 to prevent empty initial state.
  2. Added force-loading logic for task session messages with debouncing.
  3. Updated session-actions to return full assistant text and immediately persist native messages.
  4. Fixed caching logic in instance-shell2 to retain active task sessions in memory.

  Gemini AI · 2025-12-27 20:36:43 +04:00
  1e991d9ebd
• fix: flush session persistence on chat unmount to prevent data loss on navigation
  Gemini AI · 2025-12-27 13:10:39 +04:00
  5022a23aeb
• fix: load messages on mount to prevent empty chat and sync valid task persistence types
  Gemini AI · 2025-12-27 13:02:35 +04:00
  802e3ba759
• fix: persistence data loss and better auth error handling
  Gemini AI · 2025-12-27 12:55:57 +04:00
  326825a5fb
• fix(ui): export flushSessionPersistence from sessions barrel file
  Gemini AI · 2025-12-27 12:46:06 +04:00
  875be6d903
• fix(ui): flush session persistence on close to prevent data loss
  Gemini AI · 2025-12-27 12:41:23 +04:00
  9fb237e7b8
• feat(ui): add back button to session view and auto-clear active session on instance stop
  Gemini AI · 2025-12-27 12:36:25 +04:00
  c5b62f31a8
• feat: implement manual SDK session sync and fix UI crash
  Gemini AI · 2025-12-27 12:11:08 +04:00
  0e5059fc88
• Add auto-sync of SDK sessions: directly read from OpenCode storage and import to Native mode
  Gemini AI · 2025-12-27 11:46:42 +04:00
  251fad85b1
• Improve session migration: scan ALL localStorage caches to handle workspace ID changes
  Gemini AI · 2025-12-27 11:35:40 +04:00
  eca090d360
• Fix session persistence: deterministic workspace IDs based on folder path, improve Antigravity auth error handling
  Gemini AI · 2025-12-27 11:31:25 +04:00
  6b87882d8d
• Fix session migration: cache SDK sessions to localStorage for auto-import on native mode startup
  Gemini AI · 2025-12-27 11:20:56 +04:00
  6fe7367eff
• Add automatic session migration when switching from SDK to Native mode
  Gemini AI · 2025-12-27 11:13:43 +04:00
  64c7fb8d47
• Fix TypeScript compilation errors in native mode session management
  Gemini AI · 2025-12-27 11:02:08 +04:00
  eaf93e2924
• Fix Native Mode Sessions: implemented fork, revert, and sync for native sessions
  Gemini AI · 2025-12-27 10:40:45 +04:00
  ad76ade6ab
• feat: Add Google Device Authorization Flow for Antigravity native mode

  - Implemented proper OAuth device flow using gcloud CLI client ID
  - Added /api/antigravity/device-auth/start endpoint
  - Added /api/antigravity/device-auth/poll endpoint with polling
  - Added /api/antigravity/device-auth/refresh for token renewal
  - Updated AntigravitySettings UI with user code display
  - Auto-opens Google sign-in page and polls for completion
  - Seamless authentication experience matching SDK mode

  Gemini AI · 2025-12-27 04:19:53 +04:00
  610057c058
• fix: Replace broken web OAuth with manual token entry for Antigravity

  - Removed Google OAuth popup flow (invalid_client error)
  - Added manual access token input option
  - Added instructions to use SDK mode with antigravity plugin (recommended)
  - Added copy button for CLI command

  Gemini AI · 2025-12-27 04:14:51 +04:00
  d98a1fc2d0
• feat: Add Antigravity provider integration + fix native mode startup

  - Added Antigravity AI provider with Google OAuth authentication
  - New integration client (antigravity.ts) with automatic endpoint fallback
  - API routes for /api/antigravity/* (models, auth-status, test, chat)
  - AntigravitySettings.tsx for Advanced Settings panel
  - Updated session-api.ts and session-actions.ts for provider routing
  - Updated opencode.jsonc with Antigravity plugin and 11 models:
    - Gemini 3 Pro Low/High, Gemini 3 Flash
    - Claude Sonnet 4.5 (+ thinking variants)
    - Claude Opus 4.5 (+ thinking variants)
    - GPT-OSS 120B Medium
  
  - Fixed native mode startup error (was trying to launch __nomadarch_native__ as binary)
  - Native mode workspaces now skip binary launch and are immediately ready

  Gemini AI · 2025-12-27 04:01:38 +04:00
  bb1c0d81f2
• fix: MCP connection - Added explicit connectAll() method with timeout - Updated mcp-connect endpoint to use connectAll - Enhanced UI feedback for connection errors
  Gemini AI · 2025-12-26 11:05:11 +04:00
  4aa4795d4b
• fix: Update Launch-Dev-Windows.bat with same fixes
  Gemini AI · 2025-12-26 10:35:07 +04:00
  c6152cb3ea
• fix: Fixed Windows launcher and installer batch scripts - Use goto for OpenCode check to avoid nested if/else issues - Replace deprecated wmic with PowerShell for arch detection - Consistent use of delayed expansion
  Gemini AI · 2025-12-26 10:21:42 +04:00
  9980225be2
• v0.5.0: Binary-Free Mode - No OpenCode binary required

  ✨ Major Features:
  - Native session management without OpenCode binary
  - Provider routing: OpenCode Zen (free), Qwen OAuth, Z.AI
  - Streaming chat with tool execution loop
  - Mode detection API (/api/meta/mode)
  - MCP integration fix (resolved infinite loading)
  - NomadArch Native option in UI with comparison info
  
  🆓 Free Models (No API Key):
  - GPT-5 Nano (400K context)
  - Grok Code Fast 1 (256K context)
  - GLM-4.7 (205K context)
  - Doubao Seed Code (256K context)
  - Big Pickle (200K context)
  
  📦 New Files:
  - session-store.ts: Native session persistence
  - native-sessions.ts: REST API for sessions
  - lite-mode.ts: UI mode detection client
  - native-sessions.ts (UI): SolidJS store
  
  🔧 Updated:
  - All installers: Optional binary download
  - All launchers: Mode detection display
  - Binary selector: Added NomadArch Native option
  - README: Binary-Free Mode documentation

  Gemini AI · 2025-12-26 02:12:42 +04:00
  4bd2893864
• Optimize message handling to prevent UI freezing
  Gemini AI · 2025-12-24 23:24:22 +04:00
  8dddf4d0cf
• Fix initialization order: Define filteredMessageIds before lastAssistantIndex

  Fixed 'Cannot access filteredMessageIds before initialization' error by
  reordering the declarations. Since lastAssistantIndex depends on
  filteredMessageIds, it must be defined after it.

  Gemini AI · 2025-12-24 22:54:20 +04:00
  fb33e2ebe0
• Fix UI freeze: Optimize reactive memos and remove trigger loops

  Critical performance fixes for MULTIX chat mode:
  
  1. isAgentThinking - Simplified to only check last message
     - Previously iterated ALL messages with .some() on every store update
     - Each getMessage() call created a reactive subscription
     - Now only checks the last message (O(1) instead of O(n))
  
  2. lastAssistantIndex - Memoized with createMemo
     - Changed from function to createMemo for proper caching
     - Added early exit optimization for common case
  
  3. Auto-scroll effect - Removed isAgentThinking dependency
     - The thinking-based scroll was firing on every reactive update
     - Now only triggers on message count changes
     - Streaming scroll is handled by the interval-based effect
  
  These combined fixes prevent the cascading reactive loop that
  was freezing the UI during message send.

  Gemini AI · 2025-12-24 22:49:31 +04:00
  6c63bb7d7d
• Fix UI freeze: Remove high-frequency logging and throttle scroll effects

  Performance optimizations to prevent UI freeze during streaming:
  
  1. message-block-list.tsx:
     - Removed createEffect that logged on every messageIds change
     - Removed unused logger import (was causing IPC overload)
  
  2. multi-task-chat.tsx:
     - Changed filteredMessageIds from function to createMemo for proper memoization
     - Throttled auto-scroll effect to only trigger when message COUNT changes
     - Previously it fired on every reactive store update during streaming
  
  These changes prevent excessive re-renders and IPC calls during message streaming.

  Gemini AI · 2025-12-24 22:27:27 +04:00
  961c8743d3
• Integrate Context-Engine RAG service for enhanced LLM responses

  Backend:
  - Created context-engine/client.ts - HTTP client for Context-Engine API
  - Created context-engine/service.ts - Lifecycle management of Context-Engine sidecar
  - Created context-engine/index.ts - Module exports
  - Created server/routes/context-engine.ts - API endpoints for status/health/query
  
  Integration:
  - workspaces/manager.ts: Trigger indexing when workspace becomes ready (non-blocking)
  - index.ts: Initialize ContextEngineService on server start (lazy mode)
  - ollama-cloud.ts: Inject RAG context into chat requests when available
  
  Frontend:
  - model-selector.tsx: Added Context-Engine status indicator
    - Green dot = Ready (RAG enabled)
    - Blue pulsing dot = Indexing
    - Red dot = Error
    - Hidden when Context-Engine not running
  
  All operations are non-blocking with graceful fallback when Context-Engine is unavailable.

  Gemini AI · 2025-12-24 22:20:13 +04:00
  743d0367e2
• Fix UI freeze by adding yield to SSE streaming loop

  Added microtask yield (setTimeout 0) after processing each batch of SSE
  lines. This allows the main thread event loop to process UI updates and
  user interaction between streaming updates, preventing the UI from
  becoming completely unresponsive during rapid streaming.

  Gemini AI · 2025-12-24 22:08:04 +04:00
  e17e7cd32e
• Add server-side timeout handling to Ollama Cloud streaming

  - Added 60 second timeout per chunk in parseStreamingResponse
  - Added 120 second timeout to makeRequest with AbortController
  - This prevents the server from hanging indefinitely on slow/unresponsive API
  
  This should fix the UI freeze when sending messages to Ollama Cloud models.

  Gemini AI · 2025-12-24 21:53:00 +04:00
  2cb9fea514
• Fix TypeScript errors in mcp-manager and instance-shell2
  Gemini AI · 2025-12-24 21:46:13 +04:00
  fd6c563f71
• Add custom agent creator, Zread MCP, fix model change context continuity

  Features added:
  - Custom Agent Creator dialog with AI generation support (up to 30k chars)
  - Plus button next to agent selector to create new agents
  - Zread MCP Server from Z.AI in marketplace (remote HTTP config)
  - Extended MCP config types to support remote/http/sse servers
  
  Bug fixes:
  - Filter SDK Z.AI/GLM providers to ensure our custom routing with full message history
  - This fixes the issue where changing models mid-chat lost conversationcontext

  Gemini AI · 2025-12-24 21:34:16 +04:00
  3501c20471
• Backup before continuing from Codex 5.2 session - User storage, compaction suggestions, streaming improvements
  Gemini AI · 2025-12-24 21:27:05 +04:00
  e8c38b0add