During January 2025, Arantius focused on improving the stability and accuracy of CWTV metadata extraction within the harry0703/yt-dlp repository. He addressed a bug in the CWTV extractor by refining how video information is fetched through API integration and enhancing error parsing logic. Using Python and leveraging skills in data extraction and web scraping, he improved the identification of series and episode numbers to ensure reliable content matching. Arantius also updated the test suite to reflect these corrections, ensuring regression safety and data correctness. This work enhanced metadata quality, reduced identification failures, and supported more accurate downstream content processing.