invoice-generator

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it ingests untrusted user data and possesses document generation capabilities.
  • Ingestion points: The skill requires inputs for client_info and line_items in SKILL.md, which are sourced from external user-provided descriptions.
  • Boundary markers: There are no explicit delimiters or instructions to ignore embedded instructions provided to separate the user-supplied data from the agent's task logic.
  • Capability inventory: The skill utilizes the office-mcp server tools including create_docx, fill_docx_template, and docx_to_pdf for document creation and manipulation (SKILL.md).
  • Sanitization: No explicit validation, filtering, or escaping of user input is defined before the data is processed into the final document artifacts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 05:18 AM