Skip to main content
Back to challenges
Challenge

Purdue Model Multi-Agent Enterprise OT/IT Architecture Team with CrewAI

Legacy OT networks violate Purdue Model segmentation, exposing critical SCADA zones to IT network breaches. Orchestrate a multi-agent team using CrewAI with defined roles (Network Auditor, Security Engineer, and Data Architect) to evaluate network traffic logs and enforce strict level-based data flow boundaries. Achieve zero policy violation false negatives across multi-zone telemetry logs.

EngineeringHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Deploy a CrewAI collaborative agent team to audit OT/IT network boundaries and enforce Purdue Model data segmentation guidelines.

Delivery guide

How work is evaluated

Evaluation

Evaluates CrewAI team output for accuracy in detecting illegal cross-level communications.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

zero_false_negatives

Ensures all direct Level 1 to Level 4 illegal connections are detected.

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

audit_recall

Recall metric for detecting true boundary violations. • target: 1 • range: 0-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Configure CrewAI Agents with distinct personas: Purdue Auditor, Firewall Architect, and Data Pipeline Lead

  • Analyze cross-level traffic (Purdue Level 0/1 to Level 4) for policy infractions

  • Formulate DMZ (Level 3.5) data relay recommendations for cloud telemetry push

Resources and assets

Reference links and supporting material

Dataset notes

Synthesized industrial netflow logs annotated with source/destination Purdue Model zones.

Challenge setup

Evaluates CrewAI team output for accuracy in detecting illegal cross-level communications.

Challenge input

JSON array of network flow tuples (src_ip, src_level, dst_ip, dst_level, protocol)

CrewAI

Role-playing multi-agent task collaboration framework

Evaluated output

JSON report detailing severe_violations, recommended_dmz_rules, and risk_score

Evaluation checks
  • Ensures all direct Level 1 to Level 4 illegal connections are detected.
  • Recall metric for detecting true boundary violations. • target: 1 • range: 0-1
Configured evaluation evidence
  • Benchmark: OT Network Security Audit
  • Audit Recall target: 1
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • CrewAI · Required
  • crewAI · Optional

Evaluation contract

  • zero_false_negatives · Weight 1
  • audit_recall · Weight 1

Metrics

  • audit_recall · Target 1

Benchmarks

  • OT Network Security Audit
  • security

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON array of network flow tuples (src_ip, src_level, dst_ip, dst_level, protocol)

    Output: JSON report detailing severe_violations, recommended_dmz_rules, and risk_score

    Execution harness: Python

Evaluation tools

  • CrewAI · framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Purdue Model Multi-Agent Enterprise OT/IT Architecture Team with CrewAI