Bud Buddy is a web application that leverages AI to automatically detect, analyze, and fix errors in your applications. The app connects to your applications logs to fetch error logs, pulls your code from GitHub, analyzes the issue, and creates GitHub issues with proposed fixes.
- 🔄 Automatic error detection from AWS CloudWatch logs
- 📂 Code analysis from your GitHub repositories
- 🤖 AI-powered (Portia) error diagnosis and fix generation
- 🚀 Automated GitHub issue creation with detailed fix proposals
- 📊 Dashboard to track and manage your code repositories
- 🔍 Interactive UI for reviewing AI-suggested fixes
- Docker and Docker Compose
- Node.js and Yarn
- Github account and personal access token
- AWS credentials and CloudWatch access
- Portia and Anthropic API keys
First, add the necessary .env files for the external dependencies for the frontend and backend
/frontend/.env
SUPABASE_URL=
SUPABASE_ANON_KEY=
AUTH_TRUST_HOST=true
AUTH_SECRET=
GITHUB_USERNAME=
GITHUB_TOKEN=
/backend/.env
export PORTIA_API_KEY=
export ANTHROPIC_API_KEY=
export AWS_ACCESS_KEY=
export AWS_SECRET=
export AWS_REGION=
export GITHUB_TOKEN=
export GITHUB_USERNAME=
export GITHUB_REPO=
Install the frontend dependencies:
cd frontend
yarn install
Build and start the application using Docker:
docker-compose up --build
- 📡 Log Collection: Bug Buddy connects to AWS CloudWatch and monitors your logs for errors
- 🧩 Code Analysis: When an error is detected, Bug Buddy fetches the relevant code from your GitHub repository
- 🧠 AI Analysis (Portia): The AI agent analyzes the error and code to determine the root cause
- 📝 Fix Generation: The AI generates a potential fix for the issue
- 🎫 Issue Creation: Bug Buddy creates a GitHub issue with details about the error and the proposed fix
- 👀 Review & Implement: You can review the issue and implement the suggested fix
- Connect your AWS and Github account Clicking 'Sign in with GitHub' will redirect you to a GitHub page requesting you to authorize Bug Buddy to make actions to GitHub on your behalf.
- Configure: Setup your repositories and log groups to monitor
- Analyze: Start the analysis process with a single click
- Review: Examine the AI-generated diagnosis and proposed fixes
- Approve: Choose to implement the suggested fix
- Issue Creation: The Issue will be created in Github
- Pull Request Creation: The Pull Request will be created in GitHub
Frontend: Astro + React + Supabase
Backend: Python + FastAPI
Containerization: Docker
AI Agent: Portia
Contributions are welcome! Please feel free to submit a Pull Request.
If you have any questions or feedback, please open an issue on this repository.