model-cost-compare

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides instructions for task classification and cost comparison based on hypothetical pricing data.
  • [PROMPT_INJECTION]: The skill processes untrusted user task descriptions (Ingestion point: SKILL.md). Boundary markers: Absent. Capability inventory: Text output only. Sanitization: Absent. The risk of indirect prompt injection is evaluated as safe due to the lack of exploitable capabilities and minimal agent influence.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 02:45 AM
Security Audit — agent-trust-hub — model-cost-compare