How Bearer CLI works

At a high level, Bearer CLI scans the files in a project, creates an AST representation of the code using tree sitter, performs data discovery and classifications and updates the tree, and generates an internal detectors report that each report type can use. For example, the security report checks that report against a set of rules.

When you run a scan for the first time with the default settings, the application flow looks like the following diagram:

%%{ init: { 'flowchart': { 'curve': 'stepAfter' } } }%%
flowchart TB
    direction TB
    loadrule(Load rules) --> evalrule
    scan(Start scan) --> parse(Enumerate and\n parse files)
    parse --> AST
    AST --> DNC
    DNC --> df(Generate underlying \ndetection report)
    df --> evalrule(Match and evaluate \nrules)
    evalrule --> E(Generate report)
    subgraph AST[Generate AST]
    direction LR
    AST1(Tree sitter) --> |Processes source| AST2(AST)

    subgraph DNC[Detection Engine]
    direction LR
    dd(Detect data types) --> cd(Classify data types)

With the exception of loading rules from the bearer-rules repo and checking the latest version, everything happens locally in your environment.

While there is some variation for each report type, the files to AST to detection report process is the core of each Bearer CLI scan. Learn more by reviewing explanations of the sensitive data flow process and dynamic severity system.

Want to know more about inner workings of Bearer CLI? Let us know on the community Discord or on GitHub.

Ready to take the next step? Join the Bearer Cloud waitlist.