Meeting Competitive Intel

Pass

Audited by Gen Agent Trust Hub on Mar 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: A comprehensive analysis of the skill's logic and data gathering patterns indicates no security risks. The skill operates using established platform capabilities for a productivity use case.
  • [PROMPT_INJECTION]: This skill possesses an indirect prompt injection surface because it processes untrusted content from meeting transcripts. While participants in a meeting could theoretically speak instructions intended to manipulate the agent, the risk is mitigated by the agent's limited internal capabilities.
  • Ingestion points: Data is ingested from transcripts via the get_meetings_for_period action and various /api/insights/ endpoints.
  • Boundary markers: The skill does not explicitly define delimiters to separate transcript data from its internal instructions.
  • Capability inventory: Capabilities are limited to calendar and crm data retrieval and internal semantic search tools.
  • Sanitization: The skill description does not specify any sanitization or filtering of transcript text before analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 3, 2026, 11:37 AM