Sorry for the late reply.
I'm not sure that I understand your first question. Are you saying that the input that you provide does not work? If that is the case, can you provide the error (for example upload the output file)?
Just an FYI regarding numerical differentiation. If analytical gradients are available, as is the case for MP2 through the CC module, then you can use those by specifying:
where the two values specify the numerical and analytical differentation order, respectively.
Regarding the order of sections. By section you mean the double star (**), or do you also include single star (*) and dot (.)? The double stars are main sections, single stars are subsections, and dots are keywords. Thus, a dot must be below the appropriate double or single star, and single star must be below the appropriate double star. Within those the order does not matter, e.g., you can provide "**WAVE FUNCTION" before "**INTEGRALS", except that the file must begin and end with "DALTON INPUT" and "END OF DALTON INPUT", respectively. The general structure of the input file is described in section 4.2.4 in the manual.
1) Do I really need **INTEGRALS section for those jobs?
Yes, in this case, which is for the CC module, but it is not always necessary to provide the keywords of all needed integrals in other cases.
2) Is there any existing tool to update existing molecule spec with results of optimization run ? I can write a script myself, but I'd prefer existing solution.
I believe cclib (https://cclib.github.io/
) may be able to do it but I'm not sure. Some of us are working on a Python platform that will make this easier (https://aip.scitation.org/doi/10.1063/1.5144298
). However, it is still in an early development stage.