When you name something (see NameIt), it's good to have a definition for it. A definition for a term is a proposition which is true only in interpretations where the term denotes things which meet the definer's notion of what the thing might be. The definition helps keep you from changing the meaning of the term, and if you share it, it lets others with similar knowledge reuse the name and/or equate it with another name for the same thing.
So in the worst case, using the NameIt example, Alice's definition of "Charlie" is simply a statement of everything Alice knows about Charlie, but in the best case it can be must simpler and shorter. Here Alice could define "Charlie" as "The (only) guy I met this past weekend whose name is 'Charlie'."
Philosophically speaking, this sounds a bit like the descriptivist approach attacked by Kripke (see Wikipedia's article about WikiPedia:Saul_Kripke), but here we are contemplating an artificial language and an artificial (agent) society, which need not follow actual human practices.
You have two choices for sharing your definition: