Python subclass of Random using an RTLSDR as the entropy seed. As of now, does not provide a CSPRNG. While changing the seeding for a Mersenne-Twister (what random() uses internally) should prevent attacks, this should not be used for functions that need cryptographically secure random numbers.

Along with NIST SP800-22rev1 (statistical analysis of pseudo-random number generators), the Statistical Test Suite can be used to actually perform the tests outlined in this document. Runs under Windows, Linux, and Mac OSX. Written in C.

A method for distributing cryptographic entropy in the form of numerical values to systems that need it. Collects entropy from other systems and hashes it into the pool. While I can't vouch for how much entropy is in those values it seems interesting to experiment with.

The primary instance of the NIST Randomness Beacon. Has a REST API for access (which is going to change). Pull random numbers out of it but don't trust them because you didn't generate them yourself.