This repository demonstrates how to build AI agents that automatically categorize and tag GitHub issues using Azure AI services. The project includes multiple implementations showcasing different approaches to intelligent issue management.
What This Project Does
This project provides practical implementations of AI agents that can automatically analyze GitHub issues and apply appropriate tags based on content analysis using various Azure AI technologies.
Key Features
- Automated Issue Tagging: AI-powered analysis of GitHub issue content for automatic tag assignment
- Azure Integration: Built using Azure AI Projects SDK and Azure OpenAI services
- Real-time Processing: Continuous monitoring and processing of new GitHub issues
- Web Interface: Flask-based web application for manual issue categorization
Use Cases
Perfect for:
- Repository Management: Automatically organize GitHub issues with consistent tagging
- Developer Productivity: Reduce manual effort in issue triage and categorization
- AI Agent Development: Learn practical implementation patterns for AI automation
- Azure AI Integration: Understand how to build agents using Azure AI services
Technologies Used
- Azure AI Projects SDK: Core framework for building and deploying AI agents
- Azure OpenAI: Language model integration for content analysis
- Prompty: Framework for prompt engineering and management
- GitHub API: Integration for fetching and updating GitHub issues
- Flask: Web interface for manual interaction and testing
- Python: Primary programming language for all implementations
How It Works
The agents analyze GitHub issue titles and descriptions, then use AI models to determine appropriate tags from a predefined set. Different implementations showcase various approaches from simple rule-based tagging to sophisticated multi-agent orchestration with evaluation layers.