OnboardGPT for VS Code
🚀 AI-powered codebase onboarding with Structure Intelligence - Quickly understand any project with intelligent analysis, architectural pattern detection, and smart recommendations.
Features
- 🔍 One-Click Analysis: Run comprehensive codebase analysis with a single command
- 📊 Rich Markdown Reports: Get detailed insights in an easy-to-read format
- 🌐 Multi-Language Support: Works with JavaScript, Python, Java, Go, Rust, and more
- ⚡ Real-time Progress: Visual feedback during analysis
- 🎯 Actionable Insights: Get specific guidance on running and developing the project
- 🧠 Structure Intelligence: AI-powered architectural pattern detection and organization analysis
- 💡 Smart Recommendations: Get actionable suggestions for improving code organization
- 🏗️ Pattern Recognition: Automatically detect MVC, Clean Architecture, microservices, and more
- 🤖 AI Summaries: Generate intelligent summaries tailored to different audiences
- 🤖 AI Agent System: Revolutionary AI agents that understand what your project actually does
- 🎯 Project Purpose Analysis: AI explains what the project is, who uses it, and why
- 👥 User-Focused Insights: Prioritizes what developers need to know to get started
- 🔧 Configurable AI: Use your own HuggingFace API key for enhanced features
Getting Started
Installation
- Install the extension from the VS Code Marketplace
- Open any project folder in VS Code
- You're ready to analyze!
Using the Extension
- Open a project: File → Open Folder (select any codebase)
- Run analysis: Press
Ctrl+Shift+P
(or Cmd+Shift+P
on Mac)
- Search for: "OnboardGPT: Onboard Codebase"
- View results: A new document opens with your analysis report
![Demo GIF showing the analysis in action]
What You'll Get
📊 Project Overview
- Total file count and key files identified
- Technology stacks and package versions
- Entry points for running the application
🔧 Technology Detection
Automatically identifies:
- Package Managers: npm, pip, cargo, maven, gradle
- Frameworks: Express, React, Flask, Django, Spring Boot, and more
- Development Tools: TypeScript, Vite, Jest, pytest, etc.
🚀 Entry Point Discovery
Finds how to run the project:
- Web Applications: Servers, APIs, web frameworks
- CLI Tools: Command-line utilities and scripts
- Libraries: Modules and packages with exports
🎯 Getting Started Guide
- Language-specific installation instructions
- How to run detected applications
- Suggested next steps for development
🤖 AI Understanding: What This Project Is About
🎯 Project Purpose
This is a full-stack web application that provides real-time task management and collaboration features for teams. It combines a React frontend with an Express.js backend and uses PostgreSQL for data persistence.
👥 Who Uses This
- Development Teams: For managing sprints, tasks, and project milestones
- Project Managers: For tracking progress and resource allocation
- Stakeholders: For viewing project status and deliverables
⚙️ How It Works
The application uses a modern web architecture with a React SPA frontend communicating with a RESTful API backend. Real-time updates are handled through WebSocket connections, and the system includes authentication, role-based access control, and data visualization features.
💰 Business Value
Reduces project management overhead by 40% through automated task tracking, provides real-time visibility into project status, and integrates with existing development workflows to improve team productivity.
🧠 Structure Intelligence
📊 Organization Metrics
- Structure Score: 85%
- Directory Depth: 4 max, 2.3 avg
- Project Type: web-application (medium scale)
- Confidence: 90%
🏗️ Architectural Patterns
React Component Architecture (80% confidence)
- Component-based architecture using React for building reusable UI components
- Evidence: Found components directory, Using React framework
💡 Recommendations
✅ Excellent Repository Organization (low priority)
This web-application project demonstrates excellent organizational practices...
💡 Consider Adding Integration Tests (medium priority)
While unit tests are present, integration tests would improve confidence...
Example Analysis Report
# 🚀 OnboardGPT Analysis: react-express-app
## 📊 Overview
- **Total Files**: 847
- **Technology Stacks**: 1
- **Entry Points**: 2
## 🔧 Technology Stack
### TypeScript
**12 packages detected:**
- 📦 `express`: ^4.18.0
- 📦 `react`: ^18.2.0
- 📦 `prisma`: ^5.0.0
- 🔧 `vite (dev)`: ^5.0.0
## 🚀 Entry Points
### 🌐 Web Applications
- **`server/index.ts`**
- Routes/Functions: app, listen, apiRoutes
### 📚 Libraries & Modules
- **`src/index.ts`**
- Exports: App, components, utils
## 🧠 Structure Intelligence
### 📊 Organization Metrics
- **Structure Score**: 85%
- **Directory Depth**: 4 max, 2.3 avg
- **Project Type**: web-application (medium scale)
- **Confidence**: 90%
### 🏗️ Architectural Patterns
**React Component Architecture** (80% confidence)
- Component-based architecture using React for building reusable UI components
- Evidence: Found components directory, Using React framework
### 💡 Recommendations
✅ **Excellent Repository Organization** (low priority)
This web-application project demonstrates excellent organizational practices...
💡 **Consider Adding Integration Tests** (medium priority)
While unit tests are present, integration tests would improve confidence...
## 🎯 Getting Started
### Running the Web Application
```bash
npm install
npm run dev
Dependencies
- Node.js:
npm install
or yarn install
## Supported Languages
| Language | What's Detected | Package Managers |
|----------|----------------|------------------|
| **JavaScript/TypeScript** | React, Express, Vue, Angular | npm, yarn |
| **Python** | Flask, Django, FastAPI | pip, pipenv, poetry |
| **Java** | Spring Boot, Maven projects | maven, gradle |
| **Go** | HTTP servers, CLI tools | go modules |
| **Rust** | Web frameworks, binaries | cargo |
## Extension Commands
| Command | Description | Keyboard Shortcut |
|---------|-------------|-------------------|
| `OnboardGPT: Onboard Codebase` | Analyze current workspace | `Ctrl+Shift+P` → search |
## Settings
Configure OnboardGPT through VS Code settings:
| Setting | Description | Default |
|---------|-------------|---------|
| `onboardgpt.huggingfaceToken` | HuggingFace API token for AI features | None |
| `onboardgpt.modelName` | AI model to use | `Qwen/Qwen2.5-Coder-32B-Instruct` |
| `onboardgpt.enableAISummary` | Enable AI-generated summaries | `true` |
| `onboardgpt.enableStructureIntelligence` | Enable structure intelligence analysis | `true` |
| `onboardgpt.enableAIRecommendations` | Enable AI-powered recommendations | `true` |
| `onboardgpt.enableAIAgents` | Enable AI agent system for user-focused analysis | `true` |
| `onboardgpt.agentMaxFiles` | Maximum files for AI agents to analyze | `50` |
| `onboardgpt.agentMaxDepth` | Maximum directory depth for AI agents | `5` |
| `onboardgpt.summaryAudience` | Target audience for summaries | `developer` |
### Getting a HuggingFace API Key
1. Visit [HuggingFace](https://huggingface.co/settings/tokens)
2. Create a free account
3. Generate a new API token
4. Add it to your VS Code settings
5. Enjoy enhanced AI-powered analysis!
## Troubleshooting
### No Analysis Results?
- **Check workspace**: Make sure you have a folder open (not just individual files)
- **Verify project type**: Extension works best with projects that have package files (package.json, requirements.txt, etc.)
- **Large projects**: Analysis may take a few seconds for very large codebases
### Extension Not Working?
1. **Reload VS Code**: `Ctrl+Shift+P` → "Developer: Reload Window"
2. **Check output**: View → Output → Select "OnboardGPT" from dropdown
3. **Report issues**: File a bug report on our GitHub repository
## Privacy & Data
- **Local Analysis**: All code analysis happens locally on your machine
- **No Data Sent**: Your code never leaves your computer
- **No Telemetry**: We don't collect usage data or analytics
- **Optional AI**: AI features only work if you provide your own API key
- **Secure**: API keys are stored locally in VS Code settings
## Contributing
Found a bug or want to add support for a new language?
1. Visit our [GitHub repository](https://github.com/yourusername/onboardgpt)
2. Check existing issues or create a new one
3. Submit a pull request with your improvements
## Roadmap
- 🔄 **Real-time Updates**: Watch files and update analysis automatically
- 🌐 **Web Interface**: Browser-based analysis dashboard
- 🔗 **IDE Integration**: Support for more editors beyond VS Code
- 🎨 **Custom Templates**: Create your own analysis report templates
## Release Notes
### v1.0.0
- 🎉 Initial release
- ✅ Multi-language support (JavaScript, Python, Java, Go, Rust)
- ✅ Entry point detection
- ✅ Technology stack analysis
- ✅ Rich Markdown reports
### v1.2.0
- 🧠 Structure Intelligence with AI-powered analysis
- 🏗️ Architectural pattern detection (MVC, Clean Architecture, etc.)
- 💡 Smart recommendations for code organization
- 🤖 AI-generated summaries with HuggingFace integration
- 🔧 Configurable AI settings and model selection
- 📊 Advanced organization metrics and insights
---
**Happy coding!** 🚀
*Built with ❤️ by the OnboardGPT team*