An AI-powered DevRel content publishing system that automatically analyzes GitHub repositories and generates relevant content.
- GitHub Repository Analysis: Automatically extract commits, issues, PRs, and doc changes
- Topic Generation: AI-generated content topics based on repository activity
- Content Creation: Generate blog posts, code examples, and social media content
- Human-in-the-Loop: Interactive UI for selecting topics and editing content
- Content Storage: Save finalized content to a PostgreSQL database
The system is built using:
- Backend: Python with CrewAI and CopilotKit
- Frontend: Next.js with CopilotKit React components
- Database: PostgreSQL for content storage
- Python 3.10+
- Node.js 18+
- PostgreSQL database
- OpenAI API key
- GitHub token (optional but recommended)
-
Clone the repository
-
Setup the backend
cd agent
# Create and activate a virtual environment (recommended)
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies
poetry lock && poetry install
# Run FastAPI backend
poetry run demo- Setup the frontend
# Repo root
npm install- Configure environment variables
Create a .env file in the agent directory with the following:
# OpenAI API Key
OPENAI_API_KEY=sk-your-api-key
# GitHub API Token (optional but recommended to avoid rate limiting)
GITHUB_TOKEN=ghp_your-github-token
# PostgreSQL Database URL
POSTGRESQL_URL=postgresql://username:password@localhost:5432/heydev_db
# Server Port
PORT=8000
npm run dev- Access the application
Open your browser and navigate to:
- Frontend: http://localhost:3000
- Backend API: http://localhost:8000
- Enter a GitHub repository URL
- Select the start date for analysis
- Choose from AI-generated content topics
- Edit and refine the generated content
- Save and publish the content to your database
Run the integration tests:
cd agent
python -m unittest tests.integration