replicate

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as a routing mechanism to a specialized agent (reproducibility-auditor) and does not contain any instructions that execute code, access the network, or manage credentials.
  • [NO_CODE]: No executable scripts, command-line operations, or dependencies are included with this skill. The 'disable-model-invocation: true' setting further restricts the execution environment.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 10:10 AM
Security Audit — agent-trust-hub — replicate