Introduction

Content creators often encounter this scenario: after finishing video editing and copywriting, it only hits them at the moment of clicking “Publish” that they need to check—whether there are prohibited words, if the product efficacy mentioned in the voiceover can be written, whether AI-generated content requires labeling, or how long the authorization for reposted material lasts. The standards vary across platforms; Douyin, Xiaohongshu (Little Red Book), WeChat Video Accounts (Channels), and Kuaishou have vastly different requirements regarding advertising identifiers, traffic diversion, and sales pitches. Manual verification is slow and prone to omissions.

The Skill introduced below turns “pre-publication review” into a reusable checking workflow. It is neither a formatting tool nor a simple wrapper for a banned word list; rather, it conducts a comprehensive risk assessment covering visible content, voiceover expressions, sources and rights, and commercial attributes before you hit publish.

What is this

This is a Self-Media Publication Review Skill, located at @user_51a18fff/media-check. Published by community developer user_51a18fff (Version 1.0.0), it is sourced from the community and categorized under content-creation. Designed for major domestic self-media publishing scenarios, it organizes pre-publication checks into a fixed workflow, outputting review conclusions and platform adaptation suggestions.

Supported platforms include Douyin, Xiaohongshu, WeChat Video Accounts, and Kuaishou.

Core Features

It covers a comprehensive range of content types, classified by material form:

  1. Videos (including voiceovers, subtitles, and cover images)
  2. Images and Graphic Posts
  3. Audio
  4. Articles
  5. Live Streaming Scripts
  6. Product Descriptions related to e-commerce
  7. Reposted Materials and Derivative Works (interpretations/remixes)
  8. Overseas Content

The review dimensions are categorized by purpose:

  1. Visible Content: Text and elements appearing in visuals, copy, subtitles, and covers.
  2. Voiceover Expressions: Whether the spoken content involves risks such as exaggerated efficacy or absolute terms.
  3. AI Disclosure: Whether AI participation in generated content requires labeling.
  4. Sources and Rights: Material provenance, repost authorization, and whether permissions were obtained for derivative works.
  5. Commercial Attributes: Whether it constitutes advertising, requires ad identifiers, or involves traffic diversion.
  6. Platform Compliance: Whether expressions involving platform names are compliant.
  7. Live E-commerce: Risk points in sales pitches, product claims, and live scripts.
  8. Platform-Specific Versions: Adapting the same content according to different platforms’ publication standards.

Triggering is straightforward; simply state your intent in the conversation. Examples include: “Can I post this?”, “Review before publishing”, “Check for banned/sensitive words”, “Will this be shadowbanned?”, “Check voiceover/subtitles/cover/traffic diversion/sales”, or “Label AI/repost/derivative/ad/personal opinion”. It also supports retaining review conclusions, such as “Review based on my experience library”, “Save this review”, or “Review this content”.

A notable feature is requires_api_key=false, indicating this Skill runs without external paid API keys, suggesting the review logic is based on built-in checklists.

Installation and Activation

This Skill is distributed in the standard SKILL.md format. Activation is consistent with similar Skills: place the SKILL.md file into the skill directory readable by the Agent, and the Agent will load it on-demand when necessary.

Specific directory locations and installation commands for various clients are not provided in the captured data, so assumptions are not made here. If you are using Cursor, Codex CLI, or Claude Code, it is recommended to first confirm the skill loading path for your specific client before manually placing the file; the materials do not detail the differences between these three.

Official Page: https://www.skillhub.cn/skills/user_51a18fff/media-check

Typical Usage

Since specific official example commands are not provided in the materials, the following reproducible scenarios are listed based on the Skill’s positioning. Actual formatting should follow your client’s standards.

To check if a video can be published, provide the material and script to the Agent and specify the target platform:

I am preparing to post this video on Douyin. Please conduct a pre-publication review, focusing on the voiceover and subtitles.

To confirm if the copy contains sensitive words or traffic diversion risks:

Check this graphic post copy for banned words and traffic diversion risks. The platform is Xiaohongshu.

To handle AI-generated and reposted content labeling:

This part is AI-generated, and there is a section of reposted material. Please help confirm how to label them and if authorization proof is required.

When dealing with e-commerce and live scripts, explicitly stating the commercial nature improves review accuracy:

This is a live streaming sales script. Please review the product claims section and check if advertising identifiers are required.

To retain review conclusions, you can ask the Agent to save and re-verify against your experience library:

Review this content based on my experience library and save this review conclusion for future reference.

Applicable Scenarios & Caveats

Suitable for: Content distributed across multiple platforms simultaneously requiring a quick check of platform differences before publishing; complex material forms (mixed video, audio, graphics, and live scripts) where manual verification costs are high; and unified handling of compliance requirements like AI labeling, repost authorization, and ad identifiers.

Limitations to note:

  1. The review result is a risk warning, not an official platform audit. Final publication eligibility depends on platform rules.
  2. Banned words and platform standards change. The Skill’s built-in checklists need to be updated accordingly; do not treat them as a one-time, permanent library.
  3. Compliance standards for overseas content differ from domestic platforms. This requires analysis combined with target region regulations. The Skill provides a general checking framework but does not replace local legal judgment.
  4. requires_api_key=false means there is no external key dependency currently. However, if future versions connect to official platform APIs, configuration items may change; refer to the official page description.
  5. Installation methods are not detailed in the materials. If you are unsure how to load it, it is recommended to confirm the directory structure from the official page before proceeding.

Conclusion

The value of this Skill lies in consolidating scattered, easily missed pre-publication checkpoints into a repeatable workflow. It is especially suitable for content creators and operation teams distributing across multiple platforms. It does not alter the creation process but adds a layer of safeguarding before you click “Publish”.

Official Address: https://api.skillhub.cn/user_51a18fff/media-check
SkillHub Page: https://www.skillhub.cn/skills/user_51a18fff/media-check