What problem does it solve?
Replicating the style of successful Douyin (TikTok China) videos usually relies on vague intuition. This Skill turns reference videos into a structured, evidence-based style system: each video transcript becomes a 17-dimension DNA report, reports are aggregated into a weighted DNA document, and the document is compiled into an executable production template for new content.
Core Features & Use Cases
- Single-video DNA reports: Generate a 17-dimension analysis scaffold (hook, speech rhythm, shot language, cover frame, interaction design, etc.) from a transcript file, with statistics on sentence length, question density, person usage, and speech rate.
- Weighted aggregation: Merge multiple reports into a DNA document using per-sample weights and focus restrictions, distinguishing high-frequency patterns from outliers.
- Production template derivation: Produce a seven-part template (topic, title/cover, hook, development, turn, close, CTA) directly executable during content creation, with incremental updates as new samples arrive.
- Use Case: Collect transcripts from 10 videos of a benchmark Douyin account, run report on each, then build to obtain a DNA document and template that guide your own scriptwriting and cover generation.
Quick Start
Ask the agent to run douyin-style-profiler report on a prepared transcript file with a DNA ID and sample ID, then run build to aggregate all reports into the DNA document and production template.
