Skip to content

A classifier that can to a degree distinguish between commits that are prone to be vulnerable and ones that are not.

License

Notifications You must be signed in to change notification settings

jp-wagner/vccrosshair

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

vccrosshair

This is a CLI for a classifier I developed as part of my Bachelor Thesis. This tool can to an extent distinguish between commits that are prone to be vulnerable and ones that are not. You can find the entire thesis as well as a summary in the Thesis directory.

Table of Contents

Installation

pip3 install -r requirements.txt

Example

In this example we run the classifier over the commit that introduced the infamous heartbleed bug.

vccrosshair --repo path/to/openssl --commit 4817504d069b4c5082161b02a22116ad75f822b1
> Commit is prone to be vulnerable!
> Confidence: 0.6815684510145337
> The most significant feature was: Average added line count (per file count)

Exceptions

Vccrosshair will not work on commits that

  • do not alter any C/C++ files
  • are merge commits

Dataset

https://figshare.com/s/4dd1130c336f43f6e18c

About

A classifier that can to a degree distinguish between commits that are prone to be vulnerable and ones that are not.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages