Back to evaluations
Draft evaluation

Network Failure Root-Cause Agent with OpenAI Agents SDK and E2B Sandbox — evaluation

Evaluates root-cause accuracy and execution speed inside an isolated E2B sandbox.

Evaluation type
task based
Challenge
Network Failure Root-Cause Agent with OpenAI Agents SDK and E2B Sandbox
Difficulty
Advanced
Rigor
Not declared

The author has not specified a rigor level.

Open linked challenge

Evaluation overview

How the linked challenge is judged: tasks, benchmarks, and criteria count.

Tasks
1
Benchmarks
0
Criteria
0

Task templates

Inputs and expected outputs.

Task 1

network_switch_diag_task

Parses OT network syslogs and control signals to identify root switch failure.

Input format

ZIP archive containing network PCAP logs and PLC alarm records

Output format

JSON report specifying failed_device_ip, failure_timestamp, and impacted_loops