Detect disguised toxicity and harmful intent in real time

I kept running into the same issue while building community platforms: moderation tools either missed obvious toxicity or were too aggressive and killed normal conversations.

Keyword filters don’t catch things like "id1ot" or "kys", and most AI tools don’t really understand context or intent.

After dealing with this repeatedly, I built SentinelText to solve that specific gap — detecting disguised toxicity and explaining why something is harmful, not just assigning a score.

It started as something I needed myself, and now I’m exploring if others have the same problem.