israeli-personal-assistant

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill explicitly instructs the agent to fetch and use data from the public HebCal API (see scripts/morning-brief.py and the SKILL.md troubleshooting/reference sections referencing https://www.hebcal.com), and that external holiday/date data is parsed and used to drive scheduling decisions (short-day rules, deadline adjustments), so third‑party content can materially influence agent actions.