Added Oct 26, 2026
Yuwen Publish Pre-check
Before you post to Douyin, Xiaohongshu or WeChat Channels, have AI check which lines may cause trouble and how to fix them; it remembers past mistakes.
- Ecosystem
- Agent Skills
- Skill category
- Writing
- Creator / organization
- 余温(GitHub:yuwen-cool,X:@gkxspace)
What it is
This is an open-source Chinese Skill by Yuwen for checking a post before it goes up on Douyin, Xiaohongshu or WeChat Channels. The README says you let AI review it first, to see whether it can be posted, which lines are a problem and how to fix them, and it remembers the mistakes you've made so it gets more accurate over time. It works in four steps: a rule-based word check, an AI judgement of meaning, a rewrite that keeps your meaning, and a final recheck. It carries 42 rules, each with a citation to the official source, and builds up local allow and block lists, examples and safe wording as you use it. The README rules out dodging review with homophones or split characters, and doesn't promise that every post will pass.
Who it's for:Chinese-language creators on Douyin, Xiaohongshu and WeChat Channels, and people running product promotion or brand accounts.
Before you start: It's free and open source; the code, docs and templates are under the MIT licence, while the laws and platform rules it quotes aren't relicensed by it. It needs an AI agent that can read and write files and run Python, with the agent's own costs. The README says the local rule data is never uploaded, but the AI judgement is done by your agent, so your draft is sent according to that model's rules. It offers no professional legal assurance for medical or financial content.
Based on the project's public documentation. Not tested by AI Inspiration World.