notion-meeting-intelligence

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface because it ingests untrusted data from Notion pages to generate meeting materials.
  • Ingestion points: Data enters the agent via Notion:notion-search, Notion:notion-fetch, and Notion:notion-query-data-sources as documented in SKILL.md and multiple examples (e.g., project-decision.md).
  • Boundary markers: No specific delimiters or instructions to ignore embedded instructions within Notion content are provided in the templates.
  • Capability inventory: The skill can create and update pages (Notion:notion-create-pages, Notion:notion-update-page) and add comments (Notion:notion-create-comment).
  • Sanitization: No explicit sanitization or filtering of Notion content is mentioned before it is interpolated into new documents.
  • [EXTERNAL_DOWNLOADS]: The skill configuration and workflow reference an external tool from a well-known service.
  • Source: The official Notion Model Context Protocol (MCP) server at https://mcp.notion.com/mcp is referenced in agents/openai.yaml and SKILL.md.
  • [COMMAND_EXECUTION]: The SKILL.md documentation provides manual setup instructions requiring the user to execute CLI commands to configure the Notion integration.
  • Evidence: Instructions include codex mcp add notion --url https://mcp.notion.com/mcp and codex mcp login notion.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 01:16 AM