Journalists, campaigners or social media users often write messages that can involuntarily offend others, specially when talking about underrepresented groups or vulnerable communities. There are many models that can help to identify these potentially offensive messages, but they are not commonly known or easy to use by the general user.

This project consist of creating a plug-in or web app which integrates one or several already existing models to detect offensive or harmful language in free text, so anyone can check their text for involuntary offensive messages.