recoder-code 2.2.6 → 2.2.8

package/README.md

@@ -1,424 +1,188 @@
-# 🚀 Recoder-Code - The AI Development Platform That Grows With You
+# 🚀 Recoder-Code
 
-[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
-[![Node.js Version](https://img.shields.io/badge/node-%3E%3D%2018.0.0-brightgreen.svg)](https://nodejs.org/)
-[![OpenRouter](https://img.shields.io/badge/Powered%20by-OpenRouter-blue.svg)](https://openrouter.ai/)
-[![Docker](https://img.shields.io/badge/Docker-Optional-blue.svg)](https://docker.com/)
+> **Complete AI-powered development platform** - Chat with 32+ AI models, build projects, and automate your workflow
 
-**The only AI development platform that starts simple and scales to enterprise. From 30-second setup to full collaborative development platform.**
+[![npm version](https://img.shields.io/npm/v/recoder-code.svg)](https://www.npmjs.com/package/recoder-code)
+[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)
+[![Node.js](https://img.shields.io/badge/node-%3E%3D18.0.0-brightgreen.svg)](https://nodejs.org/)
 
-## ✨ What Makes Recoder-Code Different?
+## ✨ Features
 
-### 🎯 **Grows With You**
-- **Lite Mode**: Start with simple AI chat in 30 seconds
-- **Progressive Discovery**: Learn advanced features naturally
-- **Smart Defaults**: Auto-configures based on your environment
-- **Full Platform**: Scale to collaboration, ML training, and plugins
-
-### 🤖 **32+ AI Models Included**
-- **Free Models**: DeepSeek, Qwen, Google Gemma - no subscription needed!
-- **Premium Options**: Claude Sonnet 4, GPT-4, and more when you need them
-- **Smart Selection**: Automatically picks the best model for each task
-- **Cost Optimization**: Prioritizes free models without sacrificing quality
-
-### 🏗️ **Complete Development Platform**
-- **Real-time Collaboration**: Share projects with your team
-- **ML Model Training**: Train and merge custom models
-- **Plugin Marketplace**: Extend functionality with community plugins
-- **Docker Integration**: Seamless container management
-- **Project Management**: Built-in session tracking and documentation
+- 🤖 **32+ AI Models** - Access GPT-4, Claude, Gemini, DeepSeek, and more through OpenRouter
+- 📁 **File Operations** - Create, edit, and manage files directly from chat
+- 🔧 **Project Builder** - Build complete applications with AI assistance
+- ⚡ **Fast & Free** - Free models included, instant responses
+- 🎯 **Smart Defaults** - Works out of the box with minimal configuration
+- 🔄 **Interactive Mode** - Real-time chat interface with syntax highlighting
+- 🌐 **Multi-Platform** - Windows, macOS, and Linux support
 
 ## 🚀 Quick Start
 
-### Perfect Production-Ready User Flow
-
-**⚠️ Important:** Use `npm install -g` (not `npm i` shown on some npm pages) for global installation.
+### Installation
 
 ```bash
-# 1. Install globally (use full command for global install)
 npm install -g recoder-code
-
-# 2. Launch Recoder Code
-recoder-code
 ```
 
-**That's it!** When you run `recoder-code`, you'll see:
+### Setup (30 seconds)
 
-1. **🎨 Beautiful ASCII Logo** - Professional welcome screen
-2. **👋 Personal Setup** - Enter your name and get your free API key
-3. **🤖 Model Selection** - Choose your AI model (defaults to free DeepSeek)
-4. **🚀 Ready to Code** - Interactive AI terminal launches instantly
-
-### ✨ First-Time User Experience
-
-```
-██████╗ ███████╗ ██████╗ ██████╗ ██████╗ ███████╗██████╗ 
-██╔══██╗██╔════╝██╔════╝██╔═══██╗██╔══██╗██╔════╝██╔══██╗
-██████╔╝█████╗  ██║     ██║   ██║██║  ██║█████╗  ██████╔╝
-██╔══██╗██╔══╝  ██║     ██║   ██║██║  ██║██╔══╝  ██╔══██╗
-██║  ██║███████╗╚██████╗╚██████╔╝██████╔╝███████╗██║  ██║
-╚═╝  ╚═╝╚══════╝ ╚═════╝ ╚═════╝ ╚═════╝ ╚══════╝╚═╝  ╚═╝
-
-Welcome to Recoder Code v2.2.3
-Your AI-powered development companion
-
-👋 What's your name? John
-🔑 Enter your OpenRouter API key: [your-free-key]
-🤖 Model Selection:
-  1. DeepSeek Chat v3.1 (Free) - Recommended ✨
-  2. Claude 3.5 Sonnet - Advanced reasoning
-  3. GPT-4 Turbo - OpenAI's latest
-  4. Custom model
-
-✅ Setup Complete! Welcome aboard, John! 🎉
-Ready for AI-powered development...
+```bash
+recoder-code --setup
 ```
 
-### 🔄 Returning Users
-
-For returning users, just run `recoder-code` and you'll get:
-- Instant welcome back with your saved preferences  
-- No setup required - jumps straight to AI chat
-- Your chosen model and settings remembered
-
-### 🆓 Free API Key Setup
-
-1. Visit [OpenRouter.ai](https://openrouter.ai/keys) 
-2. Sign up for free account
-3. Generate your API key
-4. Paste when prompted during setup
+You'll need a free API key from [OpenRouter](https://openrouter.ai/keys) - takes 30 seconds to get!
 
-**Default Model:** `deepseek/deepseek-chat-v3.1:free` - Completely free, no subscription needed!
-
-### All Available Commands
-
-```bash
-# 💬 AI CHAT & CODE GENERATION
-recoder-code "Build me a todo app"      # Natural language development
-recoder-code --stream "Help me debug"   # Streaming responses
-recoder-code --model claude-3.5 "..."   # Choose specific model
-
-# 🤝 COLLABORATION & WORKSPACES  
-recoder-code workspace create "My Project"  # Create team workspace
-recoder-code workspace share "My Project"   # Share for collaboration
-recoder-code collab create "Team Session"   # Start collaboration session
-recoder-code collab join session-123        # Join shared session
-
-# 🐙 GITHUB INTEGRATION
-recoder-code github list                # List open issues
-recoder-code github pr "Fix bug"        # Create pull request
-recoder-code github clone <repo-url>    # Clone repository
-recoder-code github info owner/repo     # Get repository info
-
-# 🤖 ML TRAINING & MODEL MANAGEMENT
-recoder-code train create "My Model"    # Create training session
-recoder-code train start session-123   # Start model training
-recoder-code train status              # Check training status
-
-# 🔌 PLUGIN SYSTEM
-recoder-code plugins search "formatter" # Search for plugins
-recoder-code plugins install <name>     # Install plugin
-recoder-code plugins registry start    # Start plugin registry
-
-# 🏗️ INFRASTRUCTURE & DEPLOYMENT
-recoder-code monitor setup             # Set up Prometheus monitoring  
-recoder-code infra plan               # Plan Terraform deployment
-recoder-code k8s deploy              # Deploy to Kubernetes
-recoder-code services start         # Start Docker services
-
-# 📜 SCRIPTS & AUTOMATION
-recoder-code scripts list            # List available scripts
-recoder-code scripts run setup       # Run automation script
-
-# 🛠️ SYSTEM & CONFIGURATION
-recoder-code --help                  # Show help system
-recoder-code --setup                # Interactive setup wizard
-recoder-code --info                 # Show system information
-
-### First Time Here? Try These Commands
+### Usage
 
+**Interactive Mode** (recommended):
 ```bash
-# Simple AI chat
-recoder-code "Hello world"
-
-# Build a complete project
-recoder-code "Create a React todo app with TypeScript"
-
-# Get help and see what's possible
-recoder-code --help
-
-# Check your progress and unlock features
-recoder-code --progress
+recoder-code
 ```
 
-## 📖 Feature Comparison
-
-| Feature | Lite Mode | Full Platform | Claude Code |
-|---------|-----------|---------------|-------------|
-| **Setup Time** | 30 seconds | 5 minutes | 30 seconds |
-| **AI Models** | 32+ models | 32+ models | Anthropic only |
-| **Cost** | Free models available | Free + premium | $20/month |
-| **File Operations** | ✅ Full support | ✅ Full support | ✅ Full support |
-| **Project Building** | ✅ Complete projects | ✅ Complete projects | ✅ Complete projects |
-| **Collaboration** | ❌ Single user | ✅ Real-time teams | ❌ Single user |
-| **ML Training** | ❌ Not available | ✅ Custom models | ❌ Not available |
-| **Plugin System** | ❌ Not available | ✅ Marketplace | ❌ Not available |
-| **Docker Management** | ❌ Manual only | ✅ Auto-managed | ❌ Not available |
-
-## 🎯 Advanced Features
-
-### 🔧 When You're Ready for More
-
+**One-shot Commands**:
 ```bash
-# Enable full platform (Docker required)
-recoder-code --enable-services
-
-# Start real-time collaboration
-recoder-code collab create "My Project"
-
-# Train custom models
-recoder-code train merge config.json
+# Simple query
+recoder-code "create a hello world in python"
 
-# Install community plugins
-recoder-code plugins search "linter"
+# Build a project
+recoder-code "build a REST API with Express.js"
 
-# Manage Docker services
-recoder-code services status
+# Use specific model
+recoder-code --model anthropic/claude-3.5-sonnet "analyze this codebase"
 ```
 
-### 💡 Progressive Feature Discovery
+## 📚 Examples
 
-Recoder-Code automatically suggests new features as you become ready:
-
-- **After 3 file operations**: Suggests streaming for faster responses
-- **After building projects**: Suggests collaboration features  
-- **When exploring models**: Shows ML training capabilities
-- **Based on usage patterns**: Personalized feature recommendations
-
-## 🛠️ Installation & Setup Options
-
-> **⚠️ NPM Command Important Notice**  
-> Always use `npm install -g recoder-code` (full command) for global installation.  
-> The `npm i` shorthand shown on some npm websites is incorrect for global installs and may cause issues.
-
-### Option 1: NPM Global Install (Recommended)
+### Create a Web App
 ```bash
-# Use full command for global install (not npm i)
-npm install -g recoder-code
-
-# Launch with onboarding
-recoder-code
+recoder-code "create a todo app with React and Tailwind CSS"
 ```
 
-### Option 2: From Source
+### Analyze Code
 ```bash
-git clone https://github.com/caelum0x/recoder-code.git
-cd recoder-code
-npm install
-npm link
-recoder-code --setup
+recoder-code "review my authentication system and suggest improvements"
 ```
 
-### Option 3: Docker (Services Included)
+### Debug Issues
 ```bash
-git clone https://github.com/caelum0x/recoder-code.git
-cd recoder-code
-docker-compose up -d
-npm install -g .
-recoder-code --setup
+recoder-code "why is my async function not working in app.js?"
 ```
 
-## 📋 Commands Reference
+## 🎯 Key Commands
 
-### Basic Commands
 ```bash
-recoder-code                      # Interactive chat mode
-recoder-code "your prompt"        # One-shot command
-recoder-code --help               # Context-aware help
-recoder-code --setup              # First-time setup
-recoder-code --progress           # Check your progress level
+recoder-code              # Start interactive chat
+recoder-code --setup      # First-time configuration
+recoder-code --help       # Show all options
+recoder-code --models     # List available AI models
+recoder-code --stats      # View usage statistics
 ```
 
-### Model & Performance
-```bash
-recoder-code --models             # List all available models
-recoder-code --stream "prompt"    # Streaming responses (faster)
-recoder-code --fast "prompt"      # Optimize for speed
-recoder-code --cheap "prompt"     # Optimize for cost
-```
+### Interactive Mode Commands
 
-### Project Management
-```bash
-recoder-code --init [name]        # Initialize project
-recoder-code --status             # Project status
-recoder-code --sessions           # List saved sessions
-recoder-code --resume <id>        # Resume session
-```
-
-### Services & Collaboration
-```bash
-recoder-code services status      # Check service health
-recoder-code services start       # Start all services
-recoder-code --enable-services    # Enable full platform
-recoder-code collab create "name" # Start collaboration
-```
+Once in interactive mode:
+- `/help` - Show available commands
+- `/model` - Switch AI models
+- `/clear` - Clear conversation
+- `/export` - Export chat history
+- `/quit` - Exit
 
 ## 🔧 Configuration
 
-### Smart Auto-Configuration
+Your API key is stored securely in `~/.recoder-code/.env`:
 
-Recoder-Code automatically detects:
-- **User Type**: Newcomer, Developer, Full-stack, Python dev, etc.
-- **Environment**: Docker availability, Git setup, VS Code, etc.
-- **Capabilities**: Memory, cores, available tools
-- **Project Context**: Framework, language, existing configs
+```bash
+OPENROUTER_API_KEY=sk-or-v1-...
+```
 
-### Manual Configuration (Optional)
+### Advanced Options
 
 ```bash
-# View current configuration
-recoder-code config show
+# Use faster model
+recoder-code --fast "quick question"
 
-# Set specific values
-recoder-code config set services.ml.enabled true
-recoder-code config set cli.theme dark
+# Use cheapest model
+recoder-code --cheap "simple task"
 
-# Edit configuration file
-recoder-code config edit
+# Enable streaming
+recoder-code --stream "long response"
+
+# Debug mode
+RECODER_DEBUG=true recoder-code "test query"
 ```
 
-Configuration is stored in `~/.recoder-code/config.json`
+## 🌟 Why Recoder-Code?
 
-## 🐳 Docker Services (Optional)
+- ✅ **No Configuration Hell** - Works immediately after setup
+- ✅ **Multiple Models** - Switch between 32+ AI models instantly
+- ✅ **Free Tier Available** - Use free models without a credit card
+- ✅ **File Operations** - AI can create and edit files directly
+- ✅ **Session History** - Conversations are automatically saved
+- ✅ **Fallback Support** - Auto-switches models if primary fails
+- ✅ **Production Ready** - Used in real development workflows
 
-### Automatic Service Management
+## 📖 Documentation
 
-```bash
-# Generate Docker Compose file
-recoder-code services compose
+- [Model Selection Guide](./MODEL_SELECTION.md)
+- [Interactive Guide](./INTERACTIVE_GUIDE.md)
+- [Production Deployment](./PRODUCTION_GUIDE.md)
+- [Contributing](./CONTRIBUTING.md)
 
-# Smart start (only what you need)
-recoder-code services smart-start collaboration
+## 🐛 Troubleshooting
 
-# Full platform
-recoder-code services start
+### "Command not found"
+```bash
+npm install -g recoder-code
+```
 
-# Check health
-recoder-code services health
+### "API Key Required"
+```bash
+recoder-code --setup
 ```
 
-### Services Architecture
+### "Module not found"
+```bash
+npm install -g recoder-code --force
+```
 
-- **ML Training Service** (Port 8000): Custom model training and merging
-- **Collaboration Service** (Port 3003): Real-time collaborative editing
-- **Plugin Registry** (Port 3000): Plugin discovery and management
-- **Redis** (Port 6379): Session and cache management
-- **PostgreSQL** (Port 5432): User data and project storage
+## 💡 Pro Tips
 
-## 🔐 Security & Privacy
+1. **Use Free Models First** - DeepSeek and others are free and fast
+2. **Enable Streaming** - Get responses faster with `--stream`
+3. **Save Sessions** - Your chats are auto-saved in `~/.recoder-code/`
+4. **Keyboard Shortcuts** - Use `Ctrl+C` to stop, `Ctrl+D` to exit
+5. **Debug Mode** - Set `RECODER_DEBUG=true` for detailed logs
 
-### API Key Management
-- Keys stored securely in `~/.recoder-code/.env`
-- Never logged or transmitted except to chosen AI providers
-- Support for multiple provider keys
-- Easy key rotation and management
+## 🤝 Contributing
 
-### Data Privacy
-- **Lite Mode**: No data leaves your machine except AI API calls
-- **Full Platform**: Data stored locally by default
-- **Collaboration**: Optional encrypted sharing
-- **ML Training**: Models trained on your local hardware
+We welcome contributions! See [CONTRIBUTING.md](./CONTRIBUTING.md) for guidelines.
 
-### Security Features
-- Dependency vulnerability scanning
-- Code security analysis with AI models
-- Safe execution environments
-- No telemetry unless explicitly enabled
+```bash
+# Clone the repo
+git clone https://github.com/caelum0x/recoder-code.git
 
-## 🤝 Community & Support
+# Install dependencies
+npm install
 
-### Getting Help
-```bash
-recoder-code help <command>       # Command-specific help
-recoder-code --help --advanced    # Full platform help
-recoder-code --info               # System information
+# Run locally
+node cli/run.js "test query"
 ```
 
-### Community
-- **Website**: [recoder.xyz](https://recoder.xyz)
-- **NPM Package**: [npmjs.com/package/recoder-code](https://www.npmjs.com/package/recoder-code)
-- **GitHub**: [caelum0x/recoder-code](https://github.com/caelum0x/recoder-code) (Private Repository)
-- **Issues**: Report bugs and request features
-- **Discussions**: Share workflows and tips
-- **Plugin Development**: Extend the platform
-
-### Contributing
-We welcome contributions! Areas where help is most needed:
-- New AI model integrations
-- Plugin development
-- Documentation improvements
-- Platform testing on different environments
-
-## 📊 Performance & Benchmarks
-
-### Startup Performance
-- **Lite Mode**: ~0.12s to first response
-- **Full Platform**: ~2s with all services
-- **Memory Usage**: 50MB (lite) to 500MB (full platform)
-- **Model Switching**: Sub-second model changes
-
-### Model Performance
-- **Free Models**: DeepSeek R1 - competitive with GPT-4
-- **Speed Optimization**: Automatic throughput prioritization
-- **Cost Efficiency**: Free models handle 80% of tasks effectively
-- **Quality**: Intelligent model selection maintains output quality
-
-## 🗺️ Roadmap
-
-### Current (v2.0)
-- ✅ Lite mode with 30-second setup
-- ✅ Progressive feature discovery
-- ✅ Auto-service management
-- ✅ 32+ AI models with smart selection
-- ✅ Docker integration
-
-### Next Release (v2.1)
-- 🔄 VS Code extension
-- 🔄 Web UI for collaboration
-- 🔄 Plugin marketplace launch
-- 🔄 Advanced ML training workflows
-- 🔄 Enterprise SSO integration
-
-### Future (v3.0)
-- 🔮 Multi-language support
-- 🔮 AI-powered DevOps automation
-- 🔮 Custom model fine-tuning
-- 🔮 Advanced analytics and insights
-
-## 📄 License
-
-MIT License - see [LICENSE.md](LICENSE.md) for details.
-
-## 🙏 Acknowledgments
-
-- **OpenRouter**: For providing access to multiple AI models
-- **Anthropic**: For Claude models and inspiration from Claude Code
-- **Docker**: For containerization technology
-- **Node.js Community**: For the excellent ecosystem
-- **Contributors**: Everyone who has helped improve the platform
+## 📜 License
 
----
+MIT © [Recoder Code Team](https://github.com/caelum0x)
 
-**Ready to transform your development workflow?**
+## 🔗 Links
 
-```bash
-# Install globally (use full command, not npm i)
-npm install -g recoder-code
+- [NPM Package](https://www.npmjs.com/package/recoder-code)
+- [GitHub Repository](https://github.com/caelum0x/recoder-code)
+- [OpenRouter (Get API Key)](https://openrouter.ai)
+- [Report Issues](https://github.com/caelum0x/recoder-code/issues)
 
-# Launch with beautiful onboarding
-recoder-code
+## ⭐ Star Us!
 
-# Start building amazing things!
-```
+If you find Recoder-Code useful, please star the repo on GitHub! It helps others discover the project.
+
+---
 
-*The AI development platform that grows with you.* 🚀
+**Made with ❤️ by developers, for developers**

package/cli/interactive.js

@@ -951,9 +951,16 @@ class InteractiveRecoderChat {
   }
 
   setupEventHandlers() {
+    // Event handlers already set up in setupReadlineHandlers()
+    // Just start the prompt
+    this.rl.prompt();
+  }
+
+  // Legacy method - kept for compatibility but logic moved to setupReadlineHandlers
+  setupEventHandlersOld() {
     this.rl.on('line', async (input) => {
       const trimmedInput = input.trim();
-      
+
       // Handle multi-line mode
       if (this.multiLineMode) {
         if (trimmedInput === '```') {
@@ -962,7 +969,7 @@ class InteractiveRecoderChat {
           const fullInput = this.multiLineBuffer.join('\n');
           this.multiLineBuffer = [];
           this.updatePrompt();
-          
+
           if (fullInput.trim()) {
             await this.processInput(fullInput);
           }

package/cli/lite-mode.js

@@ -61,15 +61,18 @@ class LiteMode {
     this.isLiteMode = true;
     this.disabledFeatures = [
       'ml-training',
-      'collaboration', 
+      'collaboration',
       'plugins',
       'user-auth',
       'project-management'
     ];
 
-    console.log(chalk.cyan('🚀 Running in Lite Mode - AI Chat Only'));
-    console.log(chalk.gray('   Advanced features disabled for simplicity'));
-    console.log(chalk.gray('   Run "recoder-code --services" to enable full platform\n'));
+    // Only show in debug mode
+    if (process.env.RECODER_DEBUG === 'true') {
+      console.log(chalk.cyan('🚀 Running in Lite Mode - AI Chat Only'));
+      console.log(chalk.gray('   Advanced features disabled for simplicity'));
+      console.log(chalk.gray('   Run "recoder-code --services" to enable full platform\n'));
+    }
   }
 
   /**

package/cli/run.js

@@ -82,10 +82,17 @@ class RecoderCode {
     }
     
     if (!this.apiKey) {
-      console.error('Error: OPENROUTER_API_KEY environment variable is required');
-      console.log('Please set your OpenRouter API key:');
-      console.log('export OPENROUTER_API_KEY=your_api_key_here');
-      process.exit(1);
+      // Don't exit in interactive mode - let interactive.js handle the prompt
+      if (process.argv[1] && !process.argv[1].includes('interactive')) {
+        console.log(chalk.bold.yellow('\n🔐 API Key Required'));
+        console.log(chalk.gray('─'.repeat(50)));
+        console.log(chalk.white('Recoder-Code needs an OpenRouter API key to work.\n'));
+        console.log(chalk.cyan('🆓 Get your FREE key at: ') + chalk.magenta('https://openrouter.ai/keys'));
+        console.log(chalk.gray('\nThen run setup:'));
+        console.log(chalk.white('  recoder-code --setup\n'));
+        process.exit(1);
+      }
+      return; // Return early to allow interactive mode to handle
     }
 
     // Initialize OpenAI client with OpenRouter configuration
@@ -147,8 +154,8 @@ class RecoderCode {
     }
     
     try {
-      // Only show technical messages in debug or non-user mode
-      if (process.env.RECODER_DEBUG === 'true' || process.env.RECODER_USER_MODE !== 'true') {
+      // Only show in debug mode
+      if (process.env.RECODER_DEBUG === 'true') {
         console.log(chalk.gray('🚀 Initializing enhanced AI CLI systems...'));
       }
       
@@ -196,7 +203,9 @@ class RecoderCode {
       // Setup system integrations
       this.setupSystemIntegrations();
       
-      console.log(chalk.green('✅ Enhanced systems ready'));
+      if (process.env.RECODER_DEBUG === 'true') {
+        console.log(chalk.green('✅ Enhanced systems ready'));
+      }
       
     } catch (error) {
       console.log(chalk.yellow(`⚠️  Some advanced features may be limited: ${error.message}`));
@@ -250,8 +259,8 @@ class RecoderCode {
     const logFile = path.join(logsDir, `recoder-${new Date().toISOString().split('T')[0]}.jsonl`);
     fs.appendFileSync(logFile, JSON.stringify(logEntry) + '\n');
 
-    // Log usage info if available
-    if (response?.usage) {
+    // Log usage info if available (only in debug mode)
+    if (response?.usage && process.env.RECODER_DEBUG === 'true') {
       console.log(`💰 Tokens used: ${response.usage.prompt_tokens} prompt + ${response.usage.completion_tokens} completion = ${response.usage.total_tokens} total`);
     }
   }
@@ -295,7 +304,9 @@ class RecoderCode {
     }
 
     try {
-      console.log(`🔄 Using model: ${payload.model} ${payload.models ? `with fallbacks: ${payload.models.slice(1).join(', ')}` : ''}`);
+      if (process.env.RECODER_DEBUG === 'true') {
+        console.log(`🔄 Using model: ${payload.model} ${payload.models ? `with fallbacks: ${payload.models.slice(1).join(', ')}` : ''}`);
+      }
       
       if (streamResponse && !tools) {
         // Handle streaming response

package/config.js

@@ -1,4 +1,18 @@
-require('dotenv').config();
+const path = require('path');
+const os = require('os');
+const fs = require('fs');
+
+// Load from user config directory first, then fallback to project directory
+const userEnvPath = path.join(os.homedir(), '.recoder-code', '.env');
+const projectEnvPath = path.join(process.cwd(), '.env');
+
+if (fs.existsSync(userEnvPath)) {
+  require('dotenv').config({ path: userEnvPath });
+} else if (fs.existsSync(projectEnvPath)) {
+  require('dotenv').config({ path: projectEnvPath });
+} else {
+  require('dotenv').config();
+}
 
 module.exports = {
   maxTokens: 4096,

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "recoder-code",
-  "version": "2.2.6",
-  "description": "Complete AI-powered development platform with ML model training, plugin registry, real-time collaboration, monitoring, infrastructure automation, and enterprise deployment capabilities",
+  "version": "2.2.8",
+  "description": "🚀 AI-powered development platform - Chat with 32+ models, build projects, automate workflows. Free models included!",
   "main": "index.js",
   "scripts": {
     "start": "node cli/run.js",
@@ -31,10 +31,10 @@
     "todo": "node cli/todo-manager.js",
     "todo:list": "node cli/todo-manager.js list",
     "todo:stats": "node cli/todo-manager.js stats",
-    "production:check": "npm run test && npm run lint && npm run typecheck && npm run security:audit",
+    "production:check": "npm run lint && npm run typecheck && npm run security:audit",
     "production:bundle": "npm run production:check && npm run build:production",
     "production:deploy": "npm run production:bundle && npm publish --access public",
-    "postinstall": "node -e \"console.log('\\n🚀 Recoder-Code installed successfully!\\n\\n✅ Quick Start:\\n  recoder-code\\n\\n💡 If command not found, run setup:\\n  npx recoder-code setup:global\\n\\nOr add npm global bin to PATH manually:\\n  echo \\\"export PATH=\\\\\\\"$(npm config get prefix)/bin:$PATH\\\\\\\"\\\" >> ~/.zshrc\\n  source ~/.zshrc\\n\\n🎉 Then try: recoder-code --help\\n')\"",
+    "postinstall": "node -e \"console.log('\\\\n🚀 Recoder-Code v2.2.8 installed!\\\\n\\\\n✅ Quick Start: recoder-code\\\\n💡 Get API key: https://openrouter.ai\\\\n')\"",
     "prepare": "echo 'Package prepared for distribution'",
     "prepublishOnly": "npm run prepare && npm run security:audit"
   },
@@ -45,11 +45,11 @@
     "@octokit/rest": "^22.0.0",
     "axios": "^1.12.2",
     "chai": "^6.0.1",
-    "chalk": "^5.6.2",
+    "chalk": "^4.1.2",
     "chokidar": "^4.0.3",
     "commander": "^14.0.1",
     "dotenv": "^17.2.2",
-    "inquirer": "^12.9.6",
+    "inquirer": "^9.3.8",
     "jsdom": "^27.0.0",
     "openai": "^5.21.0",
     "openrouter-client": "^1.6.0",