Two skills recommended for reducing verbose AI prose
AI summary
A post recommends hardikpandya/stop-slop for removing common AI clichés and unnecessary wording, and petergyang/no-ai-slop for detecting and cleaning up AI-style phrasing while trying to preserve the original tone. The author also describes Codex with the “6.1 SOL HIGH” setting as excessively verbose and asks whether others use skills that make AI responses more concise.
In-depth insight
The post offers two practical tools for addressing verbosity and formulaic language in AI writing and agent output, but the supplied report contains only the author’s descriptions and no independent effectiveness verification.
- hardikpandya/stop-slop:去掉 AI 常见套话和废话,适合写作与 Agent 输出。
- petergyang/no-ai-slop:检测并清理 AI 腔,同时尽量保留原本语气。
另外大家有让 AI 继续写单测吗,每次写个代码测试构建跑的时间巨长。从那些大神的宣传来看,很多都是做集成测试或者黑盒测试了,不做单测了。