This project aims to develop a new approach for rating mental health apps, taking into account the subjective needs of users in the mental health community. In particular, our aims are to show that 1) an automatic rating of safety in relation to mental health apps is possible, and 2) such an automatic rating system can take into account subjective needs of app users. Our goal is to build a prototype of an automatic rating system for affective safety, while also demonstrating the limitations of what cannot be automated in such a rating system. The Covid-19 pandemic has driven demand for apps as key tools, either mandatory or voluntary in nature, for responding to social and scientific challenges including both mental health and public health in general. Yet we have not updated app regulation standards or rating procedures for a world where apps are often mandatory, and where people in a situation of vulnerability increasingly use apps based on business models originaging in the commercial sphere. This scaling up and the public concern over health apps during the pandemic have created an explicit need for the definition and rating of apps with regard to affective user safety. This project aims to contribute to giving apps a sounder basis in standards and regulatory clarity, and in turn to make informed trust in apps more possible.