solublempnn

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external biological data, creating a potential surface for indirect prompt injection.
  • Ingestion points: SKILL.md Step 1 (Record input structure or sequence).
  • Boundary markers: None identified in instructions.
  • Capability inventory: Step 2 and 3 involve verifying execution paths and command construction for modeling tools.
  • Sanitization: No specific data validation steps are listed.
  • [COMMAND_EXECUTION]: The workflow involves managing execution paths and saving command parameters for scientific tools, which is consistent with the skill's primary biological design purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 04:04 AM
Security Audit — agent-trust-hub — solublempnn