AI Wingman VSCode Extension

AI Wingman is a powerful VSCode extension that integrates AI capabilities directly into your development environment. It allows developers to interact with multiple AI providers, enhancing productivity and code quality.

Version

Current version: 1.6.16

License & Privacy

• License: See LICENSE for the complete End User License Agreement
• Privacy: See PRIVACY.md for our privacy policy

Getting Started

1. Install the AI Wingman extension
2. Configure your preferred AI provider:
   1. Press F1 or Ctrl+Shift+P
   2. Run AIWingman: Select AI Service
      • OpenAI
      • GitHub Copilot
      • AWS Bedrock
3. Run AIWingman: Select chat model to choose your model
4. Set up API credentials using the configuration commands

Features

• 🔄 Multi-Provider Support: Choose between OpenAI, GitHub Copilot, or AWS Bedrock
• 🤖 Flexible Model Selection: Configure models for each provider including GPT-4o and Claude models
• 💬 Interactive Chat Interface: Real-time streaming chat with syntax highlighting in dedicated sidebar
• 🔧 Context Menu Integration: Quick access to AI features from editor and explorer
• 📝 Code Operations:
  • Explain code functionality
  • Suggest improvements
  • Refactor code
  • Add error handling
  • Complete partial code
  • Generate comments/documentation
  • Find and fix problems
  • Generate unit tests
  • Translate between programming languages
• 🧠 Smart Memory Management:
  • Add/remove files and folders for context
  • Support for multiple file types including CSV, PDF, Office documents
  • Visual file decorations showing memory status
  • Memory usage display in status bar
  • Clear memory on demand
• 🔒 Secure Configuration: VSCode secret storage for API credentials
• 📊 Token Management: Real-time token counting and usage tracking
• ⚡ Performance Optimized:
  • Streaming responses
  • Efficient memory usage
  • Responsive webview UI
• 🎨 Theme Support: Adapts to VSCode theme automatically
• 💾 Chat History: Persistent chat conversations with context retention
• ✏️ Scratch Pad: Markdown notepad with multiple themes
• 🔧 Error Diagnostics Integration: Automatic error fixing suggestions

Commands

Provider Configuration

• AIWingman: Select AI Service: Choose between OpenAI, Copilot, or Bedrock
• AIWingman: Select chat model: Choose specific AI model for selected service
• AIWingman: Set API Key: Configure API key for selected service
• AIWingman: Set API Url: Set custom API endpoint
• AIWingman: Remove API Url: Remove custom API endpoint

Interface

• AIWingman: Open Chat UI: Launch chat interface in sidebar
• AIWingman: Open Scratch Pad: Open markdown notepad with theme support

Code Operations (Available in context menu and command palette)

• AIWingman: Ask questions to AIWingman: General AI assistance
• AIWingman: Explain the selected code: Get detailed code explanations
• AIWingman: Improve the selected code: Receive improvement suggestions
• AIWingman: Refactor the selected code: Get refactoring recommendations
• AIWingman: Add error handling to the selected code: Enhance error handling
• AIWingman: Complete the selected code: Complete partial code
• AIWingman: Add comments to the selected code: Generate documentation
• AIWingman: Find problems with the selected code: Analyze for issues
• AIWingman: Write Unit Tests for the selected code: Generate comprehensive tests
• AIWingman: Translate code from one language to another: Convert between languages
• AIWingman: Fix errors in the active editor: Fix issues in current file

Memory Management

• AIWingman: Add to memory: Add selected files/folders to AI context
• AIWingman: Remove from memory: Remove files/folders from context
• AIWingman: Clear Memory: Reset all context memory

Configuration

The extension provides extensive configuration options:

AI Service Settings

• aiWingman.selectedService: Choose AI provider (openai, copilot, bedrock)
• aiWingman.copilotChatModel: Selected GitHub Copilot model
• aiWingman.maxTokens: Maximum tokens for AI responses (default: 48000)
• aiWingman.model: AI model name (default: "aiwingman")

OpenAI Settings

• aiWingman.apiUrl: Custom OpenAI API endpoint
• aiWingman.apiKey: OpenAI API key (stored securely)

AWS Bedrock Settings

• aiWingman.bedrockRegion: AWS region (default: "us-east-1")
• aiWingman.bedrockAccessKey: AWS access key
• aiWingman.bedrockSecretKey: AWS secret key
• aiWingman.bedrockModelId: Bedrock model ID (default: "anthropic.claude-3-7-sonnet-20250219-v1:0")

Feature Settings

• aiWingman.enableCompletions: Enable/disable code completions

Prompt Customization

Customize prompts for each operation:

• aiWingman.promptPrefix.explain: Custom explain prompt
• aiWingman.promptPrefix.refactor: Custom refactor prompt
• aiWingman.promptPrefix.improve: Custom improve prompt
• aiWingman.promptPrefix.errorHandling: Custom error handling prompt
• aiWingman.promptPrefix.complete: Custom completion prompt
• aiWingman.promptPrefix.comment: Custom commenting prompt
• aiWingman.promptPrefix.findProblems: Custom problem finding prompt
• aiWingman.promptPrefix.unitTest: Custom unit test prompt

Markdown Notepad Settings

• markdownNotepad.theme: Notepad theme (default, github, elegant, dark)
• markdownNotepad.storageLocation: Custom storage location
• markdownNotepad.backupLocation: Custom backup location

User Interface

Sidebar Integration

The extension adds a dedicated AIWingman sidebar with:

• Interactive chat interface with streaming responses
• Syntax highlighting for code blocks
• Context-aware conversations
• Chat history persistence

Context Menu Integration

Right-click menus in:

• Editor: AI operations available when text is selected
• Explorer: Memory management for files and folders

Status Bar

• Memory usage indicator showing current context size
• Quick access to memory management

File Decorations

Visual indicators in the explorer showing which files are in AI memory context.

Security

• API credentials are stored securely using VSCode's secret storage
• No sensitive information is logged or stored in plain text
• All communication is handled through secure channels
• Memory context is cleared on demand
• Private license with restricted distribution

Performance

The extension is optimized for:

• Efficient message streaming and real-time responses
• Memory-efficient context management with file decorations
• Smart error handling and recovery
• Token usage optimization and tracking
• Responsive webview UI with proper resource management

Requirements

• VSCode version 1.96.0 or higher
• Internet connection for cloud AI providers
• Node.js for extension runtime
• Appropriate API credentials for chosen AI service

Supported File Types

The memory system supports various file types including:

• Source code files
• CSV files
• PDF documents
• Office documents (Word, Excel, PowerPoint)
• Text files and documentation

Architecture

The extension uses a modular architecture with:

• Service factory pattern for AI provider abstraction
• Memory manager for context handling
• Webview providers for UI components
• Command handlers for user interactions
• Configuration service for settings management

Support

For issues, feature requests, or questions, please contact AI Wingman support.

License

See LICENSE for the complete End User License Agreement.

Privacy

See PRIVACY.md for our privacy policy.