FIN500-Principles-of-Finance
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted user input via the 'argument-hint' parameter for financial calculations and conceptual analysis, creating an indirect prompt injection surface.
- Ingestion points: User input provided through the primary skill interface as described in SKILL.md.
- Boundary markers: Absent; there are no instructions for the agent to ignore or delimit potentially malicious instructions embedded in user-supplied finance data.
- Capability inventory: The skill primarily uses the provided markdown files as a reference; no dangerous tools such as network or file-system modification tools are explicitly referenced in the instructions.
- Sanitization: No sanitization or validation logic is defined for the processed input.
Audit Metadata