You may find this link useful
http://www.cpan.org/misc/cpan-faq.html#How_contribute_modules

As for finding an idea to make a module, I would say that a good module is the result of finding a solution for a practical need.
A good idea is not enough. Either you find something that is needed but nobody was able to make it or you invent a new need for which you also provide the solution. An example of a need that did not exist before its solution is the spreadsheet: before its invention, people did not have the notion that such a tool was needed.
There is a whole range of situations between the completely useless module not even good for a laugh, and the life changing module that saves you a week of work.
On the practical side, I don't think you will become famous by writing a module after picking a good idea from somebody.
However, if you really must, examine what you can do best, and pick the best piece of code that you have ever produced. Now ask yourself this question "would anybody want to use it?"
If the answer is yes, ask somebody else as well. Show your code around. Perlmonks is the perfect place for this. You must get a feeling about your skills. If you submit broken code to the CPAN, you might become famous in a different way. ;)
After you have discussed your code and you got some positive feedback, start packing your brilliancy into a module, and put LOTS of good documentation in it. Remember that even the best module without documentation is useless.
Good luck.

 _  _ _  _  
(_|| | |(_|><
 _|
[download]