I wish there was a site that would recognise common manipulation techniques in the text and based on that give it some sort of trustworthiness score. This could

I wish there was a site that would recognise common manipulation techniques in the text and based on that give it some sort of trustworthiness score. This could be then embedded in a search engine. It also work as a browser's plugin and highlight text that contains such patterns, so that the reader could be made aware that the publisher is trying to manipulate. You don't need to actually verify whether the content of the message is true or not, but a reader knowing that text is trying to manipulate, they could try to verify the story themselves.