meeting-briefing

Warn

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: MEDIUMPROMPT_INJECTIONDATA_EXFILTRATIONNO_CODE
Full Analysis
  • [DATA_EXFILTRATION]: The skill instructs the agent to pull information from highly sensitive sources, including corporate Email, Chat (Slack, Teams), Documents (Box, SharePoint), and CLM/CRM systems. While this is the skill's primary purpose, the access to PII and privileged legal data is extensive.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted data from external sources (Email, Chat) during the synthesis of briefings.
  • Ingestion points: SKILL.md (Calendar, Email, Chat, Documents, CLM, CRM).
  • Boundary markers: Absent; the instructions do not include delimiters or warnings to ignore embedded instructions in source data.
  • Capability inventory: Implied read access to communication and document platforms.
  • Sanitization: Absent; there are no instructions for sanitizing or validating external data.
  • [NO_CODE]: This skill consists solely of instructions and templates in a Markdown file. No executable code or scripts are present.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 7, 2026, 03:21 PM
Security Audit — agent-trust-hub — meeting-briefing