Meeting Command Center Plan
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection through data ingestion.
- Ingestion points: The skill reads meeting details, contact notes, deal history, and past meeting transcripts from CRM and calendar tools via
execute_actioncalls. - Boundary markers: The instructions do not explicitly define strict delimiters or "ignore instructions" wrappers for the external data being processed by the LLM.
- Capability inventory: The skill is granted access to the
calendarandcrmcapabilities. It does not have file-write or arbitrary network access. - Sanitization: The skill includes a specific privacy safeguard, instructing the agent to never include raw transcript excerpts in the output and to only summarize key themes.
Audit Metadata