Pollinations.ai Codex - AI Coding Assistant for VS Code

Pollinations.ai Codex is a powerful AI-driven assistant for Visual Studio Code, bringing intelligent coding features powered exclusively by Pollinations.ai. It offers a modern chat panel, project-wide context understanding, syntax-highlighted responses, and smart code insights.

✨ Features

🤖 Intelligent AI Chat Interface

• Modern sidebar chat panel with Cursor.com-style design
• Real-time conversations with AI about your code
• Project context awareness - understands your entire codebase
• Markdown rendering with syntax highlighting
• Message history within sessions
• Dynamic model selection - switch between different AI models on-the-fly

⚡ Smart Code Completion

• Inline code suggestions with ghost text
• Context-aware predictions based on project structure
• Multi-line completions for functions and blocks
• Debounced requests for optimal performance
• Intelligent caching to reduce API calls

🔧 Code Actions & Refactoring

Access powerful AI-driven actions through the lightbulb menu:

• 💡 Explain Selection - Get detailed explanations of code
• 🔧 Refactor Selection - Improve code quality and readability
• 🐛 Find Bugs - Detect potential issues and vulnerabilities
• 📚 Add Documentation - Auto-generate comments and documentation
• ✅ Generate Tests - Create comprehensive unit tests
• ⚡ Optimize Performance - Enhance code efficiency
• 🎯 Generate from Comments - Turn natural language comments into code

🎮 Command Palette Integration

Full integration with VS Code's command palette:

• Ctrl+1 - Open AI Chat panel
• Ctrl+2 - Explain selected code
• Ctrl+3 - Refactor selected code
• Ctrl+4 - Find bugs in code
• Ctrl+5 - Add documentation
• Ctrl+6 - Generate unit tests
• Ctrl+7 - Optimize performance
• Ctrl+8 - Generate code from comments

🏗️ Advanced Context Understanding

• Comprehensive project analysis - scans all open files
• Language detection and import analysis
• Automatic file tree building for project structure
• Smart prioritization of relevant files
• Caching system for improved performance

🚀 Quick Start

2. Configuration

1. Get an API key from Pollinations.ai (recommended, free tier)

2. Configure in VS Code Settings:

   {
     "pollinationsAI.apiKey": "your-api-key-here",
     "pollinationsAI.apiUrl": "https://enter.pollinations.ai/api/generate/v1",
     "pollinationsAI.model": "claude-large"
   }
   

3. Or use Settings UI: Ctrl+, → Search "pollinationsAI"

3. Start Using

• Press Ctrl+1 (Cmd+1 on Mac) to open AI Chat
• Start typing in any file to get inline completions
• Select code and press Ctrl+. for code actions
• Use keyboard shortcuts for quick access

🎯 Keyboard Shortcuts

📋 Usage Examples

Example 1: Explain Code

// 1. Select code you want to understand
const fibonacci = (n) => n <= 1 ? n : fibonacci(n-1) + fibonacci(n-2);

// 2. Press Ctrl+2 or click "Explain" in code actions
// 3. View detailed explanation in new editor window

Example 2: Refactor Code

// 1. Select inefficient code
function slowLoop() {
  for (let i = 0; i < array.length; i++) {
    console.log(array[i]);
  }
}

// 2. Press Ctrl+3
// 3. Code is replaced with optimized version

Example 3: Generate from Comment

// Create a function that fetches user data from API

// 1. Type the comment above
// 2. Press Ctrl+8
// 3. AI generates the implementation

Example 4: Generate Tests

function add(a, b) {
  return a + b;
}

// 1. Select function
// 2. Use code action "Generate Tests"
// 3. New test file with comprehensive tests opens

Example 5: Ask AI Chat

User: "How do I optimize React components?"
AI: "Here are the best practices for React optimization:
- Use React.memo for functional components
- Implement useMemo for expensive calculations
- Lazy load components with React.lazy...
[Full detailed response with code examples]"

🏗️ Architecture

Core Components

• Extension.ts - Main entry point and service initialization
• AIService - Handles all AI operations (explain, refactor, test generation, etc.)
• ContextService - Gathers and analyzes project context
• OpenAIClient - Manages API communication with configurable models
• CursorStyleChatPanel - Modern chat interface with model selection
• EnhancedCompletionProvider - Provides inline code suggestions
• CodeActionProvider - Implements lightbulb menu actions

Supported AI Models

• Claude Large (default) - Anthropic's advanced model
• Claude - Standard Claude model
• OpenAI GPT - OpenAI's GPT series
• Gemini Search - Google's Gemini with search capabilities

⚙️ Configuration

All settings in VS Code Settings (Ctrl+,):

🛠️ Troubleshooting

API Key Issues

Problem: "API Key not configured"
Solution:
1. Ctrl+, → Search "pollinationsAI.apiKey"
2. Paste your API key
3. Restart VS Code

Completion Not Working

Problem: No inline suggestions appear
Solution:
1. Check API key is configured
2. Ensure file is saved and active
3. Try waiting 1-2 seconds before typing (cache load)
4. Check in DevTools (F12) for errors

Chat Not Responding

Problem: Chat messages not getting responses
Solution:
1. Check API key in settings
2. Check network connection
3. Monitor API quota/rate limits
4. Restart extension: Cmd+Shift+P → "Developer: Reload Window"

📄 License

MIT License - Free for personal and commercial use

🔗 Resources

• VS Code Extension API
• Pollinations.ai Discord
• Telegram

Made with ❤️ for developers. Happy coding! 🚀