email-formatting

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its formatting script.
  • Ingestion points: Data is ingested from JSON files or CLI arguments in scripts/format_email.py.
  • Boundary markers: No delimiters or isolation instructions are present to prevent untrusted data from influencing the HTML structure.
  • Capability inventory: The skill is limited to generating and printing HTML strings; it lacks network access or file system modification capabilities.
  • Sanitization: Input data is directly interpolated into HTML templates using f-strings and the .format() method without escaping, which allows for HTML and CSS injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 02:23 AM