external-cannbot-model-model-infer-prefetch

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill serves a clear, legitimate technical purpose for hardware-specific performance optimization. No malicious patterns such as obfuscation, credential theft, or unauthorized network activity were detected.
  • [COMMAND_EXECUTION]: The skill workflow involves executing profiling tools and running functional verification scripts to measure performance gains. These actions are restricted to the intended optimization and testing scope.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes untrusted data from the local environment.
  • Ingestion points: Reads model source code and external profiling data to identify optimization candidates.
  • Boundary markers: None explicitly defined to separate data from instructions during analysis.
  • Capability inventory: Includes capabilities to modify local Python files and execute scripts for performance testing.
  • Sanitization: No specific sanitization or validation of the input code or profiling data is described.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 02:39 AM
Security Audit — agent-trust-hub — external-cannbot-model-model-infer-prefetch