sec-km2-balance-sheet-distortion

Pass

Audited by Gen Agent Trust Hub on Feb 18, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • Prompt Injection (SAFE): No malicious instruction overrides or jailbreak patterns detected.
  • Data Exposure (SAFE): No hardcoded credentials or access to sensitive system files identified.
  • Remote Code Execution (SAFE): No scripts, package managers, or remote downloads are present.
  • Indirect Prompt Injection (LOW): 1. Ingestion: Processes financial balance-sheet metrics. 2. Boundary markers: Absent. 3. Capability inventory: No tool definitions or executable code provided in the skill. 4. Sanitization: Absent. Severity is LOW as the skill lacks dangerous capabilities to exploit.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 18, 2026, 09:15 AM