problem-definition
Pass
Audited by Gen Agent Trust Hub on Feb 15, 2026
Risk Level: LOWPROMPT_INJECTIONNO_CODE
Full Analysis
- NO_CODE (SAFE): The skill consists entirely of markdown instructions and does not include any scripts, binaries, or automation commands.
- INDIRECT_PROMPT_INJECTION (INFO): The skill processes untrusted user input in the form of problem definitions. Ingestion points: User-provided text prompts. Boundary markers: None. Capability inventory: None (no file-write, network, or subprocess capabilities). Sanitization: None. Because the skill lacks side-effect capabilities, the security risk from prompt injection is negligible.
Audit Metadata