form-941

Pass

Audited by Gen Agent Trust Hub on Oct 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests untrusted payroll data and uses browser automation capabilities to interact with external websites.
  • Ingestion points: Payroll data (employee names, wages, tax withholdings) collected in SKILL.md (Step 2) and filing.md (Pre-flight).
  • Boundary markers: The skill lacks explicit delimiters or instructions to treat embedded content in payroll data as untrusted.
  • Capability inventory: The skill utilizes browser automation frameworks (Playwright, Puppeteer) for web-based data entry and filing.
  • Sanitization: No specific input sanitization or validation logic is provided for the aggregated payroll data before it is processed.
  • [COMMAND_EXECUTION]: The instructions in filing.md outline workflows for using browser automation tools including Playwright, Puppeteer, Selenium, and Browserbase. These tools are intended to automate the navigation and form-filling process on authorized tax-filing provider websites (such as TaxBandits or Gusto) to complete the tax return submission.
  • [EXTERNAL_DOWNLOADS]: The skill references and fetches official tax forms, publications, and instructions from the Internal Revenue Service (IRS) and Social Security Administration (SSA) official domains and GitHub repositories.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 4, 2026, 03:19 PM