decision-matrix
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of static markdown documents and a JSON configuration for evaluation. It does not invoke shell commands or perform network operations.
- [PROMPT_INJECTION]: No evidence of malicious instructions or behavior overrides. The skill is susceptible to indirect prompt injection because it ingests user-supplied criteria and context into its decision-making logic, though the risk is negligible given its restricted capabilities. Ingestion points: user-defined decision parameters and scores in SKILL.md. Boundary markers: none. Capability inventory: text report generation. Sanitization: none.
- [DATA_EXFILTRATION]: No sensitive file paths or credentials are used.
- [REMOTE_CODE_EXECUTION]: The skill does not install packages or execute remote code.
Audit Metadata