I'm looking for a rigorous definition of the category defined (by handwaving) as such:
- objects are "logical propositions" (first order formulas?),
- morphisms are "logical deductions" between them (inference rules?),
- composition is just concatenation of deductions (so I would say in general a morphism is a proof?).
Is it there some issue regarding the size of the objects collection? Can we speak of the "class" of all formulas about e.g. sets?
For intuitionistic logic, cartesian closed categories are equivalent to what you‘re asking for. Look up the Curry-Howard-Lambek correspondence.