Safer NSFW Prompt Auditing

I have a list of AI image-generation prompts that lean into NSFW territory. For each line I need a sharp editorial eye to decide, first, whether the wording is already acceptable or if it must be rewritten to meet common safety guidelines, and second, to supply that safer alternative whenever it’s required.

Here’s what the finished work should look like for every prompt I send:

• Column A – Original prompt
• Column B – Verdict: “safe” or “rewrite”
• Column C – Your improved, safer version (only when Column B says “rewrite”)
• Column D – One-sentence rationale for the change so I can track patterns

The tone of a rewrite should preserve the creative intent but remove or soften anything that crosses typical platform rules—extreme explicitness, minors, illegal activity, or non-consensual themes. I’m not in a rush, so take the time you need to do a thoughtful pass. Consistency matters more to me than speed.

If you’ve handled moderation or safety filtering for stable-diffusion, Midjourney, DALL·E, or similar tools, let me know; that background helps.