form-1099-misc
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting untrusted external data provided by the user (payee names, TINs, and payment amounts) which is then processed to drive browser-based automation and filing workflows. This configuration presents an attack surface for indirect prompt injection if the user-supplied data contains instructions designed to manipulate the agent's actions during the filing process.\n
- Ingestion points: User prerequisites in
SKILL.mdandfiling.mdinvolve entering lists of payments, payee identities, and W-9 information.\n - Boundary markers: There are no explicit delimiters or boundary instructions defined to wrap user-provided strings when the agent interpolates them into its context or browser automation steps.\n
- Capability inventory: The skill includes instructions for browser-based filing using official IRS systems (IRIS and FIRE), which allows the agent to navigate web interfaces and submit data based on the processed inputs.\n
- Sanitization: The skill does not specify validation or sanitization routines for the data fields before they are used in the filing workflow.
Audit Metadata