douyin-viral-content

douyin-viral-content is an automated Chinese-language copywriting generator for Douyin (TikTok in China) short-video scripts. It solves the problem of consistently producing high-engagement emotional/healing-genre captions by running a fully automated pipeline the user triggers with a simple request (e.g., "generate new copy") plus source material such as a book excerpt or a memorable quote.

The workflow automatically reads historical published content to learn what performed well, calibrates strategy (element weights, view-count estimation parameters, tag choices) from that history, generates copy applying nine documented 'viral factors' (emotional power reversal, cognitive reframing, authority endorsement, fatalistic philosophy, algorithm-friendly tags, quotable lines, completion-rate optimization, interaction hooks, and BGM suggestions), then scores the draft across content quality, algorithm fit, and originality on a 300-point scale mapped to a 5-star rating. It iterates automatically until it reaches 5 stars, estimates a view-count range via a documented model (base traffic pool, quality coefficient, account weight, vertical and time coefficients, with historical bias calibration), and writes a structured Markdown file into an output folder with a slot for the user to record actual performance later, feeding the next learning cycle.

It targets Chinese social-media creators and operators in the emotional/healing niche who want a repeatable, data-calibrated content engine. It is a benign marketing content-generation skill that only reads local history files and writes Markdown output; it runs no network calls, credentials, or system-altering operations. Its viral-factor guidance does encourage vague, unverified 'authority' citations, which is a content-quality and honesty consideration rather than a security risk.