External Scripts - checkmarx-ltd/cx-flow GitHub Wiki

CxFlow is able to use inputs from external groovy scripts, according to runtime information and specific logic that is implemented in the script

For example, you can determine the Checkmarx project name that CxFlow will trigger and/or create when initiating scan and use realtime information from the webhook payload, and edit it in the desired format in the script logic.

Here is an example of how you can use a static prefix, branch name and commit hash to determine Checkmarx project name:

String cxProject = "script-prefix-" + request.getBranch() + "-" + request.getHash() 
return cxProject

The resulting project name will look like this: script-prefix-master-fa907029c049b781f961e452a375d606402102a6. For more information about the getHash() property, see the hash field documentation in ScanRequest object.

  • CxFlow will use the string returned from the script execution to determine the Checkmarx project name
  • To enable this flow add the following property to CxFlow configuration (you can use any file name):
checkmarx:
  project-script: ...\CheckProject.groovy
  • CxFlow will use the string returned from the script execution to determine the cx-team name
  • To enable this flow add the following property to cxflow configuration (you can use any file name):
checkmarx:
  team-script: ...\CheckTeam.groovy
  • CxFlow will use the boolean value returned from the script execution to determine if scan should be
  • To enable this flow add the following property to cxflow configuration (you can use any file name):
cx-flow:
  branch-script: ...\CheckBranch.groovy
  • CxFlow will use the string returned from the script execution to determine the scan comment that is added to the scan initiated by CxFlow.
  • CxFlow searches comment-script for local files only.
  • To enable this flow add the following property to CxFlow configuration (you can use any file name):
cx-flow:
  comment-script: ...\ScanComment.groovy

To filter findings, CxFlow uses configuration that looks like the following:

cx-flow:
  filter-severity:
    - Critical
    - High
  filter-category:
  filter-cwe:
  filter-status:
    - New
    - Confirmed

Now it’s possible to provide a Groovy script in the configuration. The script returns a boolean value. The value indicates if a specific finding passes the filter. For example:

cx-flow:
  filter-script: "finding.severity == 'HIGH' || (finding.severity == 'MEDIUM' && finding.status == 'URGENT')"

The new functionality was introduced to support customer’s request. it was requested more sophisticated filtering logic. For example: “return findings of High severity; in addition, return findings of Medium severity with the ‘Urgent’ status”. An example of a corresponding Groovy expression is specified above.

CxFlow passes a finding parameter to the script. The finding object represents each SAST finding that is being checked against the filter. The finding object currently has the following properties:

  • category
  • cwe
  • severity
  • status
  • state

The finding object can be easily extended to include other properties. Note that the script should compare finding properties to uppercase string values. An exception is thrown in the following cases:

  1. Both “simple” filters and a scripted filter are specified in the config.
  2. Filtering script doesn’t return a boolean value.
  3. Filtering script has invalid syntax.
  4. A runtime error happens during script execution (comparing to a non-existent property etc.)
  • CxFlow will use the string returned from the script execution to determine the JIRA project key which is added to the bug tracker and used by CxFlow to issue tickets in it
  • To enable this flow add the following property to CxFlow configuration (you can use any file name):
jira:
  project-key-script: ...\CheckProjectKey.groovy
⚠️ **GitHub.com Fallback** ⚠️