No abstract

Attack generation from an abstract model of a protocol is not an easy task. We present BIFROST (Bifrost Implements Formally Reliable prOtocols for Security and Trust), a tool that takes an abstract model of a cryptographic protocol and outputs an implementation in C of the protocol and either a proof in ProVerif that the protocol is safe or an implementation of the attack found. We use FS2PV, KaRaMeL, ProVerif and a dedicated parser to analyze the attack traces produced by ProVerif. If an attack is found then BIFROST automatically produces C code for each honest participant and for the intruder in order to mount the attack.