nemo-mbridge-perf-moe-comm-overlap

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions focus on performance tuning for Mixture of Experts (MoE) in the Megatron Bridge framework. No prompt injection or safety bypass patterns were found.
  • [SAFE]: No sensitive file paths or credentials (API keys, tokens) are accessed or hardcoded in the skill files.
  • [SAFE]: The provided shell command for performance testing (uv run python scripts/performance/run_script.py) is a standard development practice for running local scripts with a known package manager.
  • [SAFE]: No external downloads or remote code execution patterns were identified. All URLs mentioned point to internal documentation or standard developer platforms.
  • [SAFE]: There is no evidence of obfuscated code, persistence mechanisms, or privilege escalation attempts.
  • [SAFE]: The skill includes a digital signature bundle (skill.oms.sig), which is a positive security practice for ensuring the integrity of the skill files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 04:34 PM
Security Audit — agent-trust-hub — nemo-mbridge-perf-moe-comm-overlap