Checks LLM outputs and drafts for hidden steganographic leakage before you share them. Strips invisible Unicode, normalizes text, flags anomalies, and hands back a clean copy for your approval. It never sends anything itself.

What it does

  • Checks LLM outputs
  • Drafts for hidden steganographic leakage before you share them
  • Flags anomalies

In the creator’s own words.

Open in Bot ↗