form-1099-c
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: A thorough analysis of the skill's instructions, workflows, and referenced examples revealed no evidence of malicious intent or security vulnerabilities. The skill operates within its stated purpose of providing tax reporting guidance.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a data ingestion surface where untrusted financial data and form content are processed to generate reporting plans. * Ingestion points: Untrusted 1099-C box data, user-provided asset/liability lists, and debt descriptions in SKILL.md and related examples. * Boundary markers: The skill employs a structured 10-step workflow and specific prerequisite checks to isolate and validate user-supplied data before processing. * Capability inventory: Generates structured markdown tax reporting plans and provides instructions for e-filing via external IRS or third-party tax platforms as described in filing.md. * Sanitization: Employs mathematical cross-checks (Publication 4681 Worksheet 2) and sanity validation logic to ensure the accuracy and consistency of the processed data.
Audit Metadata