AI Skill Advisor

A VS Code Extension that analyzes your software projects and intelligently recommends AI Skills, Agents, MCP Servers, Copilot Extensions, and other AI-powered capabilities to improve developer productivity.

🚀 Features

Project Analysis

• Repository Structure Scanning - Analyzes directories, file types, and project layout
• Dependency Detection - Identifies npm, pip, go.mod, Cargo.toml dependencies
• Framework Recognition - Detects React, Vue, Angular, Django, Spring, and more
• Cloud Provider Detection - Identifies AWS, Azure, GCP, and Kubernetes configurations
• CI/CD Integration - Detects GitHub Actions, GitLab CI, Azure Pipelines, Jenkins
• Git Analysis - Branch info, commit history, and repository patterns

AI-Powered Recommendations

Generates context-aware recommendations based on:

• Your project's technology stack
• Detected patterns and architecture
• Missing features (testing, documentation, monitoring)
• Industry best practices

Choose how recommendations are generated via the Mode setting:

• Copilot - Uses your existing GitHub Copilot setup in VS Code, no extra configuration needed
• Claude Code - Uses your existing Claude Code setup in VS Code, no extra configuration needed
• Cloud - Uses OpenAI GPT models via your own API key (optional, for users without Copilot or Claude Code)

Skill Categories

• Development - Code quality, refactoring, productivity tools
• Documentation - API docs, README generators, docs automation
• Security - Vulnerability scanners, secret detection, security tools
• DevOps - CI/CD, monitoring, deployment automation
• Cloud - Cost optimization, infrastructure management
• AI & Agents - AI integration, agents, ML tools
• Data - SQL generation, database optimization
• Testing - Unit tests, E2E tests, testing automation

Modern WebView UI

• Dark theme inspired by GitHub Copilot, Linear, and Raycast
• Category navigation with skill counts
• Search and filtering capabilities
• Impact and effort scoring
• One-click installation

Installation Methods

1. Command Execution - Run npm/pip install commands
2. File Scaffolding - Create configuration and template files
3. GitHub Workflows - Generate CI/CD workflow files
4. Azure Pipelines - Create Azure DevOps pipeline definitions

📦 Installation

From VSIX

code --install-extension ai-skill-advisor-1.0.0.vsix

From Source

# Install dependencies
npm install

# Build the extension
npm run build

# Package the extension
npm run package

⚙️ Configuration

VS Code Settings

Set the Mode to copilot, claudeCode, or cloud depending on how you want recommendations generated. The openAiApiKey and openAiModel settings are only needed when using cloud mode.

{
  "aiSkillAdvisor.mode": "copilot",
  "aiSkillAdvisor.openAiApiKey": "your-api-key",
  "aiSkillAdvisor.openAiModel": "gpt-4-turbo-preview",
  "aiSkillAdvisor.maxRecommendations": 20,
  "aiSkillAdvisor.autoAnalyze": false,
  "aiSkillAdvisor.showStatusBar": true,
  "aiSkillAdvisor.analysisDepth": "normal",
  "aiSkillAdvisor.excludedPaths": ["node_modules", ".git", "dist"]
}

Note: openAiApiKey and openAiModel are only required if mode is set to cloud. If you already have GitHub Copilot or Claude Code configured in VS Code, just select the matching mode from the extension's settings and it works automatically — no API key needed.

Environment Variables (Cloud mode only)

# Set OpenAI API key via environment variable
export OPENAI_API_KEY=sk-your-key-here

🎯 Usage

Commands

Activity Bar

Click the lightbulb icon in the Activity Bar to access:

• Skill category tree view
• Quick navigation to recommendations
• Installation status

Status Bar

The status bar shows:

• Analysis progress
• Current status (Ready, Analyzing, Error)
• Quick access to the dashboard

🏗️ Architecture

ai-skill-advisor/
├── extension/                    # VS Code Extension (Node.js/TypeScript)
│   ├── src/
│   │   ├── extension.ts         # Main entry point
│   │   ├── services/
│   │   │   ├── projectAnalyzer.ts    # Analyzes project structure
│   │   │   ├── aiRecommendation.ts   # AI-powered recommendations
│   │   │   └── skillInstaller.ts     # Installs recommended skills
│   │   ├── views/
│   │   │   └── webview.ts        # WebView communication
│   │   └── types/
│   │       └── index.ts         # TypeScript definitions
│   └── package.json
├── webview/                      # React Frontend (Vite + React)
│   ├── src/
│   │   ├── App.tsx             # Main app component
│   │   ├── components/         # React components
│   │   └── styles/             # CSS styles
│   └── package.json
└── package.json                 # Root workspace config

🔒 Security

• No API Key Storage - API keys stored in VS Code secure storage (only applicable in Cloud mode)
• Workspace Trust - Respects VS Code workspace trust settings
• Command Confirmation - All commands require user confirmation
• Copilot / Claude Code Modes - Work using your existing local setup, no API key leaves your machine
• Gitignore Respect - Respects .gitignore patterns

🧪 Testing

# Run extension tests
npm run test

# Run in watch mode
npm run test -- --watch

📋 Skill JSON Format

{
  "id": "jest-testing",
  "name": "Jest Testing Agent",
  "category": "testing",
  "description": "AI-powered unit testing...",
  "benefits": ["Automatic test generation", "Coverage analysis"],
  "effort": "Low",
  "impact": "High",
  "confidence": 0.95,
  "installCommand": "npm install --save-dev jest @types/jest ts-jest",
  "estimatedSetupTime": "10 minutes",
  "reasoning": "Your project would benefit from automated testing..."
}

📄 License

MIT License - see LICENSE file for details.

🙏 Acknowledgments

• Inspired by Cursor and Linear
• Built with VS Code Extension API
• React + Vite for the WebView
• Powered by GitHub Copilot, Claude Code, or OpenAI, depending on your selected mode