I plan to be working on this soon. It should be able to share pieces of the script which cross references dlls and will create an API database (plain text file for now). Before I get started I have a couple of questions.

Do we intend to gather all this information together and keep it? In a database, in the application database, in bug reports, etc. Or is it more for a developer trying to get a specific application working?

What information should exist in the API database? Currently it outputs the spec file, API name, and whether it is a stub, forward, stdcall, etc. Does anyone think more information is needed?

Would it be helpful if pedump were supported so you don't need dumpbin from Visual C++ to be able to use the scripts?