form-2555
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data such as financial figures and travel history to generate tax documentation and interact with filing sites via browser automation.
- Ingestion points: User-provided inputs for personal identifiers, income, and travel dates described in SKILL.md prerequisites.
- Boundary markers: None identified in the prompt templates or instructions.
- Capability inventory: Text generation for Form 2555 drafts and browser automation for tax filing as described in filing.md.
- Sanitization: No specific input validation or sanitization logic is provided beyond instructions for the agent to review content manually.
Audit Metadata