helion-jagged-and-autotuning

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides technical instructions and code examples for using the Helion library's API (e.g., hl.jagged_tile, helion.Config). All instructions are consistent with the stated purpose of improving kernel performance and managing configurations.
  • [SAFE]: References to environment variables (e.g., HELION_CACHE_DIR, HELION_REMOTE_CACHE_BACKEND) and configuration loading are standard practices for deep learning frameworks and do not involve unauthorized access or data exfiltration.
  • [SAFE]: No evidence of prompt injection, malicious command execution, or obfuscation techniques was found in the documentation or code snippets.
  • [SAFE]: The AOT (Ahead-of-Time) workflow described is a legitimate optimization process for GPU kernels.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 08:13 PM
Security Audit — agent-trust-hub — helion-jagged-and-autotuning