event-room-intelligence
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from registration CSV exports. Fields such as names, companies, and titles are used to trigger profile lookups and are interpolated into the generated PDF and sales 'plays'.
- Ingestion points: Registration export (CSV) in
SKILL.mdPhase 1. - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified for the ingested CSV data.
- Capability inventory: The agent uses tools for live CRM access ({{CRM}}), LinkedIn scraping ({{PROFILE_LOOKUP}}), and local file writing (PDF generation).
- Sanitization: The skill performs basic title hygiene using regex but lacks comprehensive sanitization or validation of all external inputs.
- [NO_CODE]: The skill package contains only Markdown instructions and reference files. It does not include any standalone scripts or executables, relying instead on the agent's internal capabilities and specified tools.
Audit Metadata