Just wanna make sure are IRFs in dynare computed based on (positive) one standard deviation shock? i.e. say the monetary policy shock is actually a contraction shock (which brings down output, consumption etc.) in the IRF plots reported by dynare?
Dynare always takes a one standard deviation shock by default. If you shock enters with a plus sign, e.g.
this will be a positive shock (however, in a Taylor rule, this would have the interpretation of a contractionary, often called “negative”, shock). If you want a negative shock, write