notion-meeting-intelligence

Pass

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill package consists of documentation, templates, and evaluation datasets. No executable scripts, binary files, or hidden commands were found.\n- [NO_CODE]: This is an instruction-based skill that defines agent behavior through natural language prompts and templates rather than executable code. It leverages existing tools like Notion:notion-fetch and Notion:notion-search for functionality.\n- [SAFE]: The workflow includes a surface for indirect prompt injection by ingesting data via Notion:notion-fetch and having write capabilities via Notion:notion-create-pages. However, the instructions mitigate this risk by explicitly directing the agent to cite sources and distinguish between fetched Notion facts and AI-generated insights.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 26, 2026, 01:47 AM