Model-based approaches are commonly used to engineering software for safety-critical applications. Several artifacts, such as abstract specifications, formal refinements, and assertions at the implementation level require computational support for formal reasoning activities. Delivering formal verification tools that address such reasoning activities in an automated, trustable and flexible fashion remains a scientific challenge. Researchers from LORIA and UFRN have been collaborating to address this challenge and are now developing an automated theorem prover targeted at trustable software verification efforts.