RepoPilot AI helps developers understand unfamiliar GitHub repositories and discover the best way to contribute.
Built for the Microsoft AI Skills Fest Hackathon, RepoPilot AI uses Azure AI Foundry IQ to analyze repositories, explain project structure, evaluate contribution difficulty, and guide beginners toward meaningful first contributions.
Instead of overwhelming contributors with thousands of files, issues, and unfamiliar code, RepoPilot AI transforms repositories into actionable contributor intelligence.
Open source is difficult to enter.
New contributors often struggle with:
- Understanding large repositories
- Figuring out what the project actually does
- Knowing where to start contributing
- Finding beginner-friendly tasks
- Understanding folder structures
- Evaluating repository health
- Predicting maintainer responsiveness
This causes many developers to abandon open-source contributions before making their first pull request.
RepoPilot AI acts like an AI-powered open-source mentor.
Simply enter a GitHub repository URL and RepoPilot AI will:
✅ Explain the repository in beginner-friendly language
✅ Detect technologies automatically
✅ Recommend where to start contributing
✅ Analyze repository health
✅ Suggest first contribution areas
✅ Explain folder structure
✅ Provide contribution roadmaps
- AI-generated repository summaries
- Beginner-friendly explanations
- Technology stack detection
- Repository purpose breakdown
- Folder structure explanations
- Contribution roadmap generation
- Beginner difficulty scoring
- Best files/folders to start exploring
- Suggested first PR recommendations
- Contributor onboarding guidance
- Repository health score
- Maintainer activity analysis
- Issue health scoring
- Bus factor risk estimation
- Community activity indicators
RepoPilot AI uses Microsoft Azure AI Foundry IQ for grounded repository understanding.
Instead of hallucinating, the AI:
- Builds a repository knowledge base
- Uploads repository context into a vector store
- Uses Foundry File Search
- Generates grounded explanations using retrieved repository context
This produces more reliable and explainable AI outputs.
GitHub Repository
↓
GitHub REST API
↓
Repository Intelligence Engine
(Tech Stack + Metrics + Repo Analysis)
↓
Azure AI Foundry IQ
(Vector Store + File Search)
↓
Grounded AI Insights
↓
FastAPI Backend
↓
Interactive Web Dashboard
RepoPilot AI fetches:
- README content
- Repository structure
- Folder hierarchy
- Issues and contribution signals
- Maintainer activity data
- Repository metadata
The system detects:
- Programming languages
- Frameworks
- Tech stack
- Contribution areas
- Project complexity
Repository context is uploaded into a Foundry Vector Store.
Using Foundry File Search, the model retrieves grounded repository information before generating explanations.
This prevents hallucinated repository summaries.
RepoPilot AI generates:
- Beginner-friendly summaries
- Contribution roadmaps
- Folder explanations
- Difficulty assessments
- Suggested onboarding path
- Python
- FastAPI
- Uvicorn
- Microsoft Azure AI Foundry
- Azure AI Foundry IQ
- Azure Vector Store
- Foundry File Search
- GPT-4.1-mini
- GitHub REST API
- GitHub Repository Metadata
- HTML
- CSS
- Vanilla JavaScript
- Azure CLI (
az login) - DefaultAzureCredential()
Add screenshot here
Add screenshot here
Add screenshot here
Add screenshot here
RepoPilot-AI-Web/
│
├── main.py
│ ├── FastAPI application
│ └── API endpoints
│
├── github_fetcher.py
│ └── GitHub REST API client
│
├── repo_analyzer.py
│ ├── Tech stack detection
│ └── Contribution area analysis
│
├── metrics_analyzer.py
│ └── Repository health scoring
│
├── foundry_agent.py
│ ├── Azure AI Foundry integration
│ ├── Vector store creation
│ ├── File Search retrieval
│ └── Grounded AI reasoning
│
├── templates/
│ └── index.html
│
├── static/
│ ├── css/
│ │ └── style.css
│ │
│ └── js/
│ └── app.js
│
├── requirements.txt
├── .env.example
└── README.md
git clone https://github.com/your-username/RepoPilot-AI-Web.git
cd RepoPilot-AI-Web
python -m venv .venv
Activate environment:
Command Prompt (CMD)
.venv\Scripts\activate
PowerShell
.\.venv\Scripts\Activate.ps1
pip install -r requirements.txt
Copy environment template:
Windows
copy .env.example .env
Mac/Linux
cp .env.example .env
Add the following:
GITHUB_TOKEN=your_github_token
FOUNDRY_PROJECT_ENDPOINT=https://your-project.services.ai.azure.com/api/projects/your-project
FOUNDRY_MODEL_DEPLOYMENT=gpt-4.1-mini
Download Azure CLI:
Verify installation:
az --version
Login:
az login
uvicorn main:app --reload
Open browser:
http://localhost:8000
GET /api/analyze?repo_url=
GET /api/analyze?repo_url=microsoft/vscode
{
"summary": "AI-generated repository summary",
"health_score": 87,
"tech_stack": ["TypeScript", "Electron"],
"folder_explanation": {},
"contribution_guide": {},
"maintainer_activity": {},
"issue_health": {}
}
Understand unfamiliar repositories faster.
Useful for:
- GSoC
- Hacktoberfest
- First contributions
- Community onboarding
Help onboard contributors more efficiently.
Learn repository structure and technologies.
- Paste GitHub repository URL
- RepoPilot fetches repository data
- Azure AI Foundry analyzes repository context
- AI generates grounded explanations
- User receives contribution roadmap
Add demo video link here
- AI contributor match score
- Personalized issue recommendations
- PR success prediction
- Contributor skill-based onboarding
- Multi-repository comparison
- GitHub Copilot integration
- Repo learning paths
RepoPilot AI leverages Microsoft Azure AI Foundry to make open-source contribution more accessible and beginner-friendly.
By combining:
- GitHub Repository Intelligence
- Azure AI Foundry IQ
- Grounded Retrieval
- Contributor Guidance
RepoPilot AI transforms confusing repositories into actionable contributor intelligence.
Built for the Microsoft AI Skills Fest Hackathon 2026
MIT License