I could do this, however before i accept to do the project, we would need to discuss the end result, and i would like to explain to you the possibilities in detail.
I expect that you understand, that this feat can be achieved, however, usually in such cases there is no guarantee that it will work on a specific end user or for all end users.
Depending on the amount of work invested and on the budget, a specific installation percentage can be met. For example, if it is agreed to use say 5 different attack vectors, which work with some versions of IE and some versions of Firefox, but not in chrome , it can be estimated and afterwards verified by statistics, that out of 1000 site visitors, say in 200 cases the software will be silently executed.
If you understand this and accept that the quality and completion of the finished product can be measured only by whether such percentage goal is achieved, I am ready to discuss this project further.
--Also, during the course of the project, I will need confirmation by email from you, stating that you understand and consent that I provide this work for you for scientific/educational purposes only and/or security penetration testing and that you accept and understand that you are legally responsible for any legal consequences that may arise from your use of this scientific work, and that I am not and can not be held responsible for any damages, harm or losses to any person or company, should they arise, from your usage of this scientific research software. Such consent in email from from your registered email address is a requirement for the release of software, source code and documentation of the completed work.