GitHub radar
Artificial intelligence that can scan your posts before publishing them
An open-source agent skill for Claude Code, Cursor and Codex that helps you precheck your drafts based on Chinese social media regulations (Douyin, Xiaohongshu, WeChat Video). Show you the problematic keywords, give you alternatives and the corresponding laws/regulations. All the processing is done offline on your own machine.
yuwen-publish-precheck is an open-source agent skill for Claude Code, Cursor and Codex that helps you precheck your drafts based on Chinese social media regulations (Douyin, Xiaohongshu, WeChat Video). It uses a 2-step process: first, a keyword matching method to identify potential violations, then, an AI semantic analysis method to judge whether a specific keyword violates the regulations in different contexts. For example, it may be okay to say something in your life story but should not be used in business presentations. Besides, we also provide suggestions for replacing those problematic keywords. Once you replace them with the suggested alternatives, our tool will automatically check whether the modified draft still contains any violations.
Why a vibe-coder should care
You can easily install this skill by just one sentence to Claude Code, just send the github address to the agent and it will help you set up. The whole processing is done locally on your own machine without sending your drafts to any server. Moreover, we also store your previous results locally to improve the accuracy of the tool with your usage.
▌ More finds