Initial Release: OpenQode Public Alpha v1.3

docs/PROJECT_BRIEF.md

@@ -0,0 +1,91 @@
+# OpenQode v1.01 Preview Edition - Project Brief
+
+## 🎯 Overview
+
+OpenQode is a revolutionary fork that combines the power of **OpenCode's TUI interface** with **Qwen's generous free tier** and **advanced AI capabilities**. This hybrid solution delivers the best of both worlds in a single, user-friendly package.
+
+## 🚀 Key Benefits
+
+### vs Regular OpenCode
+- ✅ **2,000 free daily requests** vs paid OpenAI models
+- ✅ **No token limits** vs OpenAI token restrictions  
+- ✅ **60 RPM rate limit** for sustained productivity
+- ✅ **Automatic credential refresh** - no manual token management
+- ✅ **One-click authentication** vs complex API key setup
+
+### vs Regular Qwen Code CLI
+- ✅ **Rich TUI interface** vs command-line only
+- ✅ **Visual chat experience** vs text-based interaction
+- ✅ **Enhanced features** (Lakeview, Sequential Thinking)
+- ✅ **Windows-optimized** with batch launcher
+- ✅ **Self-contained package** vs complex installation
+
+## 🎨 Unique Features
+
+### Hybrid Architecture
+- **OpenCode TUI** for superior user experience
+- **Qwen OAuth** for seamless authentication
+- **Cross-session credential sync** for persistent access
+- **Automatic token refresh** for uninterrupted service
+
+### Enhanced AI Capabilities
+- **Lakeview Mode** - Concise, minimal output
+- **Sequential Thinking** - Structured problem-solving
+- **Multi-model support** - Coder and Vision models
+- **Context-aware responses** with advanced reasoning
+
+### User Experience
+- **One-click launch** via OpenQode.bat
+- **Smart authentication** - auto-detects and handles auth
+- **Fallback options** - manual browser opening
+- **Professional interface** with clear status indicators
+
+## 📊 Technical Advantages
+
+### Performance
+- **60 RPM rate limit** vs typical 20 RPM limits
+- **2,000 daily requests** for heavy usage
+- **Zero cost** vs paid alternatives
+- **Reliable uptime** with enterprise-grade infrastructure
+
+### Integration
+- **PowerShell-native** on Windows
+- **Cross-platform compatible** (Windows, Linux, macOS)
+- **Portable package** - no installation required
+- **Profile integration** for system-wide access
+
+## 🎯 Target Users
+
+### Developers
+- Want **free AI coding assistance**
+- Need **TUI interface** for better productivity
+- Require **high rate limits** for intensive work
+- Prefer **one-click solutions**
+
+### Teams
+- Need **cost-effective AI tools**
+- Want **consistent experience** across members
+- Require **easy deployment** and setup
+- Value **professional interfaces**
+
+## 🏆 Competitive Edge
+
+OpenQode dominates the market by offering:
+1. **Free tier** that rivals paid services
+2. **Professional TUI** missing from free tools
+3. **Enterprise features** in consumer package
+4. **Zero maintenance** with automatic updates
+
+## 🔮 Future Roadmap
+
+- **Multi-provider support** (OpenAI, Anthropic, Google)
+- **Plugin ecosystem** for extended functionality
+- **Cloud sync** for settings and sessions
+- **Team collaboration** features
+- **Advanced analytics** and usage tracking
+
+---
+
+Note: If `bin/opencode.exe` is not included in a GitHub clone, OpenQode will auto-download it during installation or first run. OpenCode UI features (for example Lakeview or Sequential Thinking) depend on the specific OpenCode build you are using.
+
+**OpenQode v1.01 Preview Edition** - Where professional AI coding meets free accessibility.

docs/README.md

@@ -0,0 +1,165 @@
+# OpenQode v1.01 Preview Edition
+
+**OpenCode + Qwen Integration Package**
+
+OpenQode is a powerful integration of OpenCode with Qwen AI models, providing free access to advanced coding capabilities.
+
+## Features
+
+### Multiple AI Models
+- Qwen Coder Model - Free 2,000 requests/day, 60 RPM
+- Qwen Vision Model - Free 2,000 requests/day, 60 RPM  
+- OpenCode Big Pickle - Default OpenCode model
+- OpenCode GPT-5 Nano - Experimental model
+- Grok Code - Grok coding model
+
+### Automatic Authentication
+- Browser-based OAuth authentication for Qwen models
+- Automatic credential management and refresh
+- One-time setup, persistent access
+
+### 🔐 Qwen Authentication
+
+When you select a Qwen model, OpenQode will automatically:
+1. Check if you're authenticated with Qwen
+2. Initiate OAuth authentication if needed
+3. Complete the OAuth flow (may happen in background)
+4. Store credentials for automatic refresh
+
+**Authentication Notes:**
+- 🌐 Browser may open automatically, or authentication may complete in background
+- 🔑 If browser doesn't open, visit: https://qwen.ai
+- 📱 Use `./scripts/qwen-auth.ps1` for manual authentication
+- 🔄 Use `./bin/opencode.exe auth logout qwen` to reset authentication
+
+**Qwen OAuth Benefits:**
+- ✅ 2,000 free requests per day
+- ✅ No token limits  
+- ✅ 60 requests per minute rate limit
+- ✅ Automatic credential refresh
+
+### Enhanced Features
+- Lakeview Mode - Concise, minimal output
+- Sequential Thinking - Structured problem-solving
+- TUI Interface - Terminal-based interaction
+
+## Quick Start
+
+### Option 1: TUI Default (Recommended)
+```powershell
+.\OpenQode.ps1
+```
+This will automatically launch the Terminal UI with Qwen Coder model.
+
+### Installation (PowerShell or Batch)
+If you cloned this repo without `bin/opencode.exe`, OpenQode will auto-download it during install or first run.
+
+```powershell
+.\Install.ps1
+```
+
+Or the batch alternative:
+```bat
+Install.bat
+REM For system-wide PATH (run as Administrator):
+Install.bat --systemwide
+```
+
+### Option 2: Model Selection Menu
+```powershell
+.\OpenQode-Menu.ps1
+```
+This will show a menu to choose your AI model.
+
+### Option 3: Direct Launch with Specific Model
+```powershell
+# Launch with Qwen Coder (default)
+.\OpenQode.ps1 -Model "qwen/coder-model"
+
+# Launch with specific model
+.\OpenQode.ps1 -Model "opencode/big-pickle"
+```
+
+### Option 4: Double-Click TUI
+Simply double-click `OpenQode.bat` in Windows Explorer to start TUI immediately.
+
+## First Time Setup
+
+1. Run OpenQode and choose a Qwen model (option 1 or 2)
+2. Your browser will open automatically
+3. Complete authentication on qwen.ai
+4. Return to terminal and press Enter
+5. Enjoy free AI coding!
+
+Note: If `bin/opencode.exe` is not present (for example, when cloning from GitHub), OpenQode will auto-download it during `Install.ps1` or on first run of `OpenQode.ps1`. You can also download it manually with `.\scripts\download-opencode.ps1`.
+
+## Model Details
+
+### Qwen Models (Free)
+- 2,000 requests per day
+- 60 requests per minute
+- No token limits
+- Automatic credential refresh
+
+### OpenCode Models
+- No authentication required
+- Standard OpenCode features
+
+## File Structure
+
+```
+OpenQode-v1.01-Preview/
+  (Note: `bin/opencode.exe` is auto-downloaded if missing)
+├── OpenQode.ps1          # Main TUI launcher (default)
+├── OpenQode-Menu.ps1     # Model selection menu
+├── OpenQode.bat          # Windows batch TUI launcher
+├── Install.ps1           # Installation script
+├── PACKAGE_INFO.txt      # Package summary
+├── bin/
+│   └── opencode.exe      # OpenCode binary
+├── scripts/
+│   ├── opencode-launcher.ps1
+│   └── opencode-interactive.ps1
+└── docs/
+    └── README.md         # This file
+```
+
+## Troubleshooting
+
+### Authentication Issues
+```powershell
+# Re-authenticate with Qwen
+.\bin\opencode.exe auth qwen
+```
+
+### Check Authentication Status
+```powershell
+# List all authenticated providers
+.\bin\opencode.exe auth list
+```
+
+### Model Switching
+Run OpenQode again and choose a different model from the menu.
+
+## System Requirements
+
+- Windows 10/11
+- PowerShell 5.1 or later
+- Internet connection for Qwen authentication
+
+## Version Information
+
+- Version: 1.01 Preview Edition
+- Release Date: December 2024
+- Components: OpenCode + Qwen Integration
+
+## Support
+
+For issues and updates, check the original repositories:
+- OpenCode: https://github.com/sst/opencode
+- Qwen Code: https://github.com/QwenLM/qwen-code
+
+---
+
+**OpenQode v1.01 Preview Edition**  
+*Powerful AI Coding, Free for Everyone*

docs/tui-redesign-plan.md

@@ -0,0 +1,120 @@
+# TUI Professional Visual Redesign - Implementation Plan
+
+## Goal
+Improve TUI text rendering to be professional-grade like Claude Code, fixing HTML entities, text wrapping, streaming jitter, and layout consistency. **VISUAL ONLY - NO functionality changes.**
+
+---
+
+## User Review Required
+
+> [!IMPORTANT]
+> **New Dependencies Required**: This plan requires installing `he` (HTML entity decoder) and `wrap-ansi` (width-aware text wrapping). Both are lightweight, pure ESM-compatible, and widely used.
+
+---
+
+## Root Cause Analysis
+
+| Issue | Source | Fix |
+|-------|--------|-----|
+| `'` showing literally | `marked` outputs HTML entities, no decoder | Add `he.decode()` before rendering |
+| Text bleeding into borders | No width calculation for wrapping | Use `wrap-ansi` with calculated width |
+| Inconsistent spacing | No theme system, hardcoded values | Create centralized theme module |
+| Streaming jitter | React re-renders on every token | Add batched streaming component |
+| List indentation broken | Inline tokens not properly wrapped | Fix list rendering in markdown |
+
+---
+
+## Proposed Changes
+
+### Phase 1: Add Dependencies
+
+```bash
+npm install he wrap-ansi
+```
+
+---
+
+### Phase 2: Create Theme Module
+
+#### [NEW] bin/tui-theme.mjs
+Central theme configuration:
+- Spacing scale: xs=0, sm=1, md=2, lg=3
+- Semantic colors: fg, muted, info, warning, error, accent, border
+- Border styles with fallback for non-unicode terminals
+- Consistent component styling
+
+---
+
+### Phase 3: Fix Markdown Renderer
+
+#### [MODIFY] bin/ink-markdown-esm.mjs
+**Changes:**
+1. Import `he` for HTML entity decoding
+2. Import `wrap-ansi` for width-aware wrapping
+3. Apply `he.decode()` to all text content before rendering
+4. Fix inline token rendering to decode entities
+5. Improve list item wrapping and indentation
+6. Add terminal width awareness for proper wrapping
+
+---
+
+### Phase 4: Update Card Components
+
+#### [MODIFY] bin/opencode-ink.mjs (lines 477-590)
+**Changes:**
+1. Import theme module
+2. Update SystemCard, UserCard, AgentCard, ErrorCard to use theme
+3. Add consistent padding/margin from theme
+4. Add flex-based width constraints
+5. Ensure proper overflow handling
+
+---
+
+### Phase 5: Add Streaming Stability (Optional)
+
+#### [MODIFY] bin/opencode-ink.mjs
+**Changes:**
+1. Add batched state update for streaming content
+2. Implement 16ms throttle to reduce jitter
+3. Use stable keys for message list
+
+---
+
+## Verification Plan
+
+### Manual Testing (User)
+Since there are no automated TUI tests in this project, verification is manual:
+
+1. **HTML Entity Test**:
+   - Launch TUI: `node bin/opencode-ink.mjs`
+   - Send message containing apostrophes: "what's the best approach?"
+   - ✅ Verify: apostrophes render as `'` not `'`
+
+2. **Text Wrapping Test**:
+   - Resize terminal to 80 columns
+   - Send long message
+   - ✅ Verify: text wraps cleanly, doesn't touch borders
+
+3. **List Rendering Test**:
+   - Ask AI to list something: "list 5 programming languages"
+   - ✅ Verify: bullet points render with proper indentation
+
+4. **Code Block Test**:
+   - Ask AI for code: "write a hello world in Python"
+   - ✅ Verify: code block has border, syntax highlighting
+
+5. **Streaming Stability Test**:
+   - Send message requiring long response
+   - ✅ Verify: output streams without major jitter
+
+6. **Functionality Preservation Test**:
+   - `/help` → Shows help card
+   - `/agents` → Shows agent menu with scrolling
+   - `/context` → Toggles context
+   - `/clear` → Clears messages
+   - ✅ All commands work exactly as before
+
+### Cross-Platform Testing
+- Windows Terminal + PowerShell
+- Windows CMD
+- Linux bash (if available)