meeting-notes

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection and Cross-Site Scripting (XSS) via processed meeting notes.\n
  • Ingestion points: Raw meeting notes provided by the user are processed as instructed in SKILL.md.\n
  • Boundary markers: There are no boundary markers or explicit instructions for the agent to ignore potentially malicious content within the meeting notes.\n
  • Capability inventory: The summarized data is passed to assets/webview.html, which serves as the presentation layer.\n
  • Sanitization: The skill lacks sanitization of the extracted data. Specifically, assets/webview.html uses innerHTML to render the 'decisions' and 'action items' fields, creating a Cross-Site Scripting (XSS) vulnerability. Mitigation: Use textContent or a secure sanitization library when rendering dynamic content.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 08:23 AM
Security Audit — agent-trust-hub — meeting-notes