- Set up project structure and environment
- Implement arXiv API integration
- Create function to search for recent AI safety papers
- Implement filtering for papers from the last few days
- Increase the number of papers fetched per day
- Update search query to include multiple relevant keywords
- Expand search to cover more categories and a longer time range
- Develop caching mechanism for fetched articles
- Set up Google Sheets integration
- Create a new sheet for storing article details
- Implement function to add new articles to the sheet
- Add informative header row to the Google Sheet
- Fix error handling in add_entry method
- Implement cleaning of old, unposted papers
- Create secrets management system
- Implement YAML-based configuration
- Create utility function for loading configuration
- Set up article download query for AI safety topics
- Integrate Anthropic API for article summarization
- Create Claude prompt for effective summarization
- Implement function to summarize articles using Claude
- Develop main script to orchestrate the backend processes
- Fetch articles, add to Google Sheets
- Implement error handling for API calls
- Implement caching mechanism
- Implement summarization
- Use the recommendations API endpoint to show recommended articles for each article we post
- For each article marked for posting, scrape <= 3 recommended articles and include them in the posts
- This can be implemented by adding a new method to the SemanticScholarAPI class and calling it here after processing each paper or in a separate function that runs after process_new_papers
- Set up Telegram bot using BotFather
- Implement bot functionality
- Create command to subscribe/unsubscribe users
- Develop function to send summaries to subscribed users
- Integrate bot with backend to receive new summaries
- Set up Twitter Developer account and create app
- Implement Twitter bot functionality
- Create function to post summaries as tweets
- Implement scheduling for regular posting
- Integrate bot with backend to receive new summaries
- Set up Hugo project structure
- Implement data fetching from Google Sheets
- Create Hugo templates for:
- Home page listing all articles
- Individual article pages with summaries
- Implement search functionality
- Set up automatic build and deployment process
- Implement error handling and logging throughout the project
- Set up unit tests for critical functions
- Create documentation for each component
- Implement rate limiting for API calls
- Set up monitoring and alerts for the system
- Implement user feedback mechanism for summaries
- Add categorization for articles
- Develop recommendation system based on user preferences