TikTok shadowban check

Give me a list of accounts and an Apify key — I return, per account, a shadowban probability, the likely causes, and what to do about it.

python3 ../../tools/shadowban_check.py handle1 handle2 --apify-key apify_api_xxx
# optional: OPENROUTER_API_KEY=... adds a "what works / what flops" LLM read

Output: one table on stdout + shadowban_reports/<date>_<handle>.md per account (≈$0.01 per account via Apify clockworks~tiktok-profile-scraper, last ≤30 posts).

The model behind the verdict

TikTok distributes in rounds. Every post gets a seed batch (~200–500 views); only if that batch's watch-time / completion / saves clear the bar does the post get the next round. So the histogram of posts by round is the diagnostic, not the average: