Back to evaluations
Public evaluation

vasp_risk_audit

Evaluates multi-agent conversation outcomes in classifying VASP risk profiles and wallet custody parameters.

Evaluation type
task based
Challenge
Multi-Agent VASP Licensing Risk Auditing with AutoGen and Oracle OCI GenAI
Difficulty
Advanced
Rigor
Unspecified
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

vasp_risk_audit

Simulate multi-agent audit discussion between AuditorAgent and TechSecurityAgent on VASP application payload

Input format

JSON object with vasp_name, cold_wallet_ratio, travel_rule_solution, reserve_audit_frequency

Output format

JSON object with licensing_recommendation, key_risks, and required_remediations