Skip to main content
Loading navigation
Back to challenges

Software-Defined Control System Anomaly Analyzer with Claude Agents SDK

Software-defined control (SDC) decouples control logic from underlying hardware, increasing exposure to network latency and switch failures. Construct an intelligent log diagnostic agent using Claude Agents SDK and Shakudo to identify control loop timing deviations and pinpoint network switch anomalies.

EngineeringHosted by Vera
Sign in free to start

What you are building

Build an SDC log diagnostic workflow leveraging Claude Agents SDK deployed on Shakudo to isolate digital network failures affecting PLC control loops.

How work is evaluated

Evaluation

Evaluates agent diagnosis accuracy when provided mixed network packet captures and virtual PLC cycle timing logs.

Datasets

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

How submissions are scored

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

root_cause_accuracy

Checks if identified component matches true failed hardware/switch.

Binary check

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

Dimension 2

diagnostic_precision

Precision of identified fault categories across 20 synthetic scenario inputs. • target: 0.95 • range: 0.8-1

Binary check

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

Learning goals

  • Configure Claude Agents SDK tools for multi-file SDC diagnostic parsing

  • Deploy model execution pipelines using the Shakudo platform runtime

  • Analyze network switch latency impact on virtual PLC cycle time stability

  • Generate actionable root-cause failure reports for automation engineers

Resources

Dataset notes

Combined virtualized PLC telemetry logs, EtherCAT packet traces, and industrial switch port traffic metrics.

Challenge setup

Evaluates agent diagnosis accuracy when provided mixed network packet captures and virtual PLC cycle timing logs.

Challenge input

JSON containing network_switch_logs and virtual_plc_cycle_times.

Claude Agents SDK

Provides sophisticated extended reasoning for complex log debugging.

Shakudo

Manages execution environment, agent scaling, and data integrations.

Evaluated output

JSON containing faulty_component, anomaly_type, and confidence_score.

Evaluation checks
  • Checks if identified component matches true failed hardware/switch.
  • Precision of identified fault categories across 20 synthetic scenario inputs. • target: 0.95 • range: 0.8-1
Configured evaluation evidence
  • Benchmark: SDC Fault Diagnostics Benchmark
  • Diagnostic Precision target: 0.95
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Action Space
  • Claude Code · Required
Observation
  • Claude Code · Required
Policy Serving
  • Claude 4.1 Opus · Optional
  • Claude 4 Sonnet · Optional

Evaluation contract

  • root_cause_accuracy · Weight 1
  • diagnostic_precision · Weight 1

Metrics

  • diagnostic_precision · Target 0.95

Benchmarks

  • SDC Fault Diagnostics Benchmark
  • security

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing network_switch_logs and virtual_plc_cycle_times.

    Output: JSON containing faulty_component, anomaly_type, and confidence_score.

    Execution harness: Python

Evaluation tools

  • Claude Agents SDK · Agentic Reasoning
  • Shakudo · Agent Infrastructure Platform

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Software-Defined Control System Anomaly Analyzer with Claude Agents SDK