nemo-relay-setup-observability

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is documentation-only, providing guidance on how to choose and configure observability outputs (Subscribers, AtifExporter, OpenTelemetry, OpenInference) for the NeMo Relay framework.
  • [SAFE]: No external dependencies are downloaded or installed. The mentioned libraries and namespaces (e.g., nemo_relay.subscribers) are internal references to the vendor's (NVIDIA) own SDK.
  • [SAFE]: No prompt injection, obfuscation, or data exfiltration patterns were detected. The skill uses standard instructional language.
  • [SAFE]: The evaluation file (evals.json) contains standard test cases used for verifying agent performance and does not introduce security risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 01:50 PM
Security Audit — agent-trust-hub — nemo-relay-setup-observability