nemo-relay-instrument-calls

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely instructional, providing documentation for developers on how to wrap existing tool and LLM calls with NeMo Relay APIs (e.g., tools.execute, llm_call_execute). it contains no executable scripts or automation that could perform unauthorized actions.
  • [EXTERNAL_DOWNLOADS]: The skill references the official GitHub repository for NeMo Relay (https://github.com/NVIDIA/NeMo-Relay/). This is an official resource from NVIDIA and is considered a safe reference.
  • [DATA_EXPOSURE]: Analysis of the skill instructions and metadata found no hardcoded credentials, API keys, or access to sensitive local file paths (e.g., .env, .ssh).
  • [COMMAND_EXECUTION]: The code patterns provided are API usage examples for Rust, Python, Node.js, and Go. There are no patterns involving os.system, subprocess, or other shell execution vectors using untrusted input.
  • [PROMPT_INJECTION]: The instructions do not contain attempts to override agent behavior or bypass safety filters. Instead, the skill focuses on implementing guardrails and lifecycle events.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 06:03 PM
Security Audit — agent-trust-hub — nemo-relay-instrument-calls