The diagram below shows ControlFlag's two main phases: firstly, the pattern mining phase, followed by the scanning for anomalous patterns. The pattern mining phase mines typical patterns in the user-provided GitHub repositories and then builds a decision-tree from the mined patterns. The scanning phase can then apply the mined patterns to flag anomalous expressions in the user-specified target repositories.

The latest version has been updated with full support for working with C code. It can now learn typical patterns and detect anomalous patterns (inference) within if statements of C programs.

The new release can be used to download GitHub repositories for C language code. It can then mine conditional expressions, and ControlFlag can be trained using the code from the repository.

It also comes with datasets that contain conditional expressions that have already been mined from GitHub repositories, and has support for detecting anomalous conditional expressions in a target repository.

As a part of scanning for anomalies, ControlFlag also suggests possible corrections in case a conditional expression is flagged as an anomaly.