Mastra AI Purdue Model Data Router with XAgent Architecture
Modern OT data architectures require strict segregation across Purdue Model levels while allowing cross-layer data synthesis. Construct a TypeScript agent workflow using Mastra AI and XAgent tooling to validate and route Level 1 SCADA tags up to Level 4 Enterprise MES systems. Success is measured by achieving zero Purdue security zone violations and 100% schema validation accuracy.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop a Mastra AI agent workflow integrated with XAgent execution patterns to securely route and format OT telemetry across Purdue Model boundaries.
How work is evaluated
Evaluates data packet validation and zone enforcement across Purdue Model level definitions.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
no_direct_level0_to_level4
Rejects packets attempting direct Level 0/1 to Level 4 transmission.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
routing_compliance_rate
Ratio of correctly validated and routed data packets. • target: 1 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Construct a Mastra AI workflow with specialized steps for Purdue Model Level 0-4 validation.
Implement XAgent tool execution patterns for automated network boundary enforcement.
Validate schema payloads transition between SCADA, Historian, and Enterprise MES.
Prevent prohibited direct cross-layer communication paths using automated policy checks.
Reference links and supporting material
Synthesized SCADA tag logs representing level-by-level telemetry hops across an industrial chemical plant environment.
Challenge setup
Evaluates data packet validation and zone enforcement across Purdue Model level definitions.
Challenge input
JSON payload containing source_level, destination_level, and telemetry_data
Mastra AI
Required framework for building TypeScript workflows and agent steps.
XAgent
Provides structured tool execution patterns for boundary checks.
Evaluated output
JSON result with allowed, route_path, and sanitization_status
- Rejects packets attempting direct Level 0/1 to Level 4 transmission.
- Ratio of correctly validated and routed data packets. • target: 1 • range: 0-1
- Benchmark: PurdueArchBench-2026
- Routing Compliance Rate target: 100%
- 1 public reference case
- JavaScript execution harness
- javascript 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
- Mastra AI · Required
- XAgent · Optional
- Mastra AI · Required
Evaluation contract
- no_direct_level0_to_level4 · Weight 1
- routing_compliance_rate · Weight 1
Metrics
- routing_compliance_rate · Target 1
Benchmarks
- PurdueArchBench-2026
- security
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON payload containing source_level, destination_level, and telemetry_data
Output: JSON result with allowed, route_path, and sanitization_status
Execution harness: JavaScript
Evaluation tools
- Mastra AI · Agent Framework
- XAgent · Agent System Tooling
Recipe freeze marker
No recipe freeze time is recorded.