Douyin Content Tracker Skill

This package largely does what it says (scrapes Douyin, extracts audio, runs Whisper), but take these precautions before installing/ running: - Confirm and set MEDIACRAWLER_DIR in the skill .env (the registry metadata omitted this required env). The skill will not work without it. - Backup your MediaCrawler installation before running this: the scripts temporarily overwrite MediaCrawler/config/base_config.py to change fetch count and then restore it. Ensure MEDIACRAWLER_DIR points to an isolated/copy location you control. - Be aware cookies: the pipeline stores Douyin session cookies in a local .douyin_cookies.json and passes them as a command-line argument to MediaCrawler. Passing secrets on a command line can expose them via process listings to other users on the same machine — avoid running this on multi-user/shared systems. After use, consider deleting or rotating the cookie. - Expect big downloads and disk usage: Playwright browser binaries and Whisper model weights (medium model ~GBs) will be downloaded; make sure you have bandwidth and storage. - Review accounts.txt and .env for any unintended targets or output directories; set OUTPUT_BASE_DIR to an isolated folder you control. - Confirm legal/ToS considerations: automated scraping may violate Douyin/TikTok terms of service or local law — ensure you have the right to scrape the targeted accounts. If you want to proceed in a safer manner: run the pipeline on an isolated VM/container or a dedicated user account, inspect the code (especially run_mediacrawler and set_mediacrawler_max_count), and avoid running the workflow on shared systems where process listing or filesystem writes could leak credentials or affect other software.