This code presents several implementations of cryptographic program obfuscation, as listed in the following table:

For some background on cryptographic program obfuscation, see https://eprint.iacr.org/2013/451.

A discussion of the AGIS implementation appears in the work:

"Implementing Cryptographic Program Obfuscation." Daniel Apon, Yan Huang, Jonathan Katz, Alex J. Malozemoff. Cryptology ePrint Archive 2014/779. https://eprint.iacr.org/2014/779.

All schemes use the graded encoding scheme of Coron et al. (CRYPTO, 2013. https://eprint.iacr.org/2013/183). The implementation is in a mix of Python and C, using Sage, GNU GMP and OpenMP. The code is under active development, and may not be stable! I will try to keep the master branch in a working state (most of the time), though.

Instructions for building and running the code are in code/README.md, and scripts for running experiments and processing the results are in the scripts/ directory. Full traces of all the experiments found in the above paper can be found in the scripts/results/ directory.

For any questions/comments, please e-mail amaloz at cs dot umd dot edu.

In order to aid in the cryptanalysis of cryptographic program obfuscation, we plan to release a series of obfuscated point functions (that is, functions that output 0 on all inputs except one), with the goal of the attacker being to learn the hidden point. The following table lists the available/broken challenges.