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