First I would recommend reading chapter 9 (Notation) of The Practice of Programming by Kernighan and Pike.
When you have done that, come back here with specific questions on how to map the concepts in that chapter to specific designs for the problems you want to solve.
The basic pattern is to write an interpreter that is passed a 'command' argument, and possibly an 'environment' argument and executes the command (in the environment). You then have the option of writing a parser, that takes a 'script' string and converts it into a valid 'command' object (ie. an external-DSL); or you provide a library to help users build the 'command' object explicitly in the same language you are using (internal-DSL).
Kernighan and Pike do a good job of showing both how trivial and how complex an interpreter can be. If you want more depth, then I would suggest reading The Essentials of Programming Languages by Daniel Friedman et al. Which builds at least one different interpreter per chapter, and demonstrates how to implement features such as variables, functions, scopes, objects, classes, static-typing, and continuations.
However I would suggest trying your hand at a trivial DSL first, otherwise it's all just theory—a book is much more interesting when it is made relevant and practical by your previous experience.