What problem does it solve? Short-drama production teams generating scripts, character prompts, and video prompts for AI platforms like Seedance 2.0 and Gemini risk having content rejected or accounts flagged for violating platform content policies. This Skill provides a structured compliance review at every production stage so violations are caught before submission. ## Core Features & Use Cases - Stage-Based Review: Checks director scripts (stage 1), character/scene/prop prompts (stage 2), and Seedance 2.0 video prompts (stage 3) against a defined acceptance checklist. - Two-Tier Policy Model: Enforces absolute red lines (politics, national security, copyrighted IP, real public figures, minors) as pass/fail gates, while treating violence density, romantic content, and religious themes as non-blocking advisory notes for overseas audiences. - Controlled Reporting: Returns a single user-readable Chinese status line to the producer while keeping detailed verdicts and advisory notes in internal records. - Use Case: Before submitting episode 1 storyboard prompts to Seedance 2.0, run the review to confirm no copyrighted character names, real celebrity references, or prohibited political content appear in the prompts. ## Quick Start Review the current episode's video prompts in 02-视频提示词.md for Seedance 2.0 and Gemini content policy compliance and report whether any red lines are touched.
