form-8824
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 property descriptions and taxpayer names, which is then utilized in a browser-based filing workflow. This creates a potential surface for indirect prompt injection attacks where malicious data could influence the agent's actions during the automated interaction with the IRS filing portal.
- Ingestion points: Prerequisites in SKILL.md (e.g., property descriptions, fair market values, and names).
- Boundary markers: The instructions do not define delimiters or explicit warnings to isolate user-supplied data from the agent's command instructions.
- Capability inventory: filing.md describes a browser automation process to interact with the IRS Free File Fillable Forms (FFFF) website.
- Sanitization: There are no instructions for sanitizing or escaping the content of user-supplied fields before they are interpolated into the filing workflow.
Audit Metadata