Here is dynamic SQL that will generate the script you need.  It will be initially displayed in a column in grid mode, left clink on the link and it will...

Or filtered indexes?

Do we need to worry about filestream data in the indexes?

Are the indexes create in non-standard location(s) such as different file groups from the tables?

Can we assume that the clustered index is not dropped?  How about indexes supporting constraints like primary or unique indexes?

Script each index while they exist in the database.  Put each in their own stored procedure.  Use a master procedure to call each individual procedure if the index is missing.

Document everything, make sure all (ALL) changes are in writing.  Keep your proposed design handy.  At some point the data in the system will hit critical mass and performance will...

If you have to keep recreating the indexes I would look for the process(es) that are dropping the indexes and fix that so the indexes don't keep disappearing.

If you work regularly with Oracle I would also push for Toad for Oracle.  It will provide tools for query performance tuning.  Not free, but if I had to work...