Turn security research into reliable pipelines.

Bring the tools, scripts, and logic behind your research into one workflow. Use the same process against the next target without rebuilding the plumbing around it.

Join betaOpen live demo
Sample workflowSuspicious dependencyRead-only preview

A real Zokor sample workflow. A Repository Push sends a Repository event to Clone Repository. Repository contents pass to Extract Dependencies, which sends a dependency collection to For Each. Each Item is analyzed by Dependency Risk Analysis and Add tag, then Tagged returns to the loop's Next Item input. When the collection is exhausted, Completed continues once to Create finding.

  1. Repository Push outputs Repository event to Clone Repository's Repository input.
  2. Clone Repository outputs Repository contents to Extract Dependencies' Repository contents input.
  3. Extract Dependencies outputs Dependencies to For Each's Collection input.
  4. For Each outputs Each Item to Dependency Risk Analysis' Current dependency input.
  5. Dependency Risk Analysis outputs Signal to Add tag's Context input.
  6. Add tag outputs Tagged back to For Each's Next Item input.
  7. For Each outputs Completed once to Create finding's Context input.
  1. 01

    Repository Push → Clone Repository → Extract Dependencies

    A GitHub push starts the workflow. The next two nodes clone the repository and collect dependencies from supported manifests and lockfiles.

  2. 02

    For Each → Dependency Risk Analysis → Add tag

    For Each sends one dependency through four checks. Add tag returns Tagged to Next Item so the loop can continue.

  3. 03

    Completed → Create finding

    After the collection is finished, Completed continues once and creates the finding.

Workflow sources

  • GitHub
  • PyPIPlanned
  • npmPlanned

The work around the research

Build the analysis, not the scaffolding.

The interesting part is getting the research working. Repeating it usually means writing more code to move data, loop through collections, branch on results, and carry context between steps.

BeforeA working analysis plus
  • scripts and scanners
  • parser and glue code
  • loops and manual handoffs
  • an output file
In ZokorThe same work represented as
  • named analysis steps
  • connections that show data movement
  • loops and branches in the graph
  • one definition for the next target