skills/malue-ai/dazee-small/onenote/Gen Agent Trust Hub

onenote

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Interacts with Microsoft OneNote via PowerShell COM objects to list, search, and create pages.
  • [PROMPT_INJECTION]: Vulnerable to indirect prompt injection through note ingestion. Ingestion points: Notebook hierarchy and page content retrieved via GetHierarchy and FindPages in SKILL.md. Boundary markers: Absent; note content is not delimited or accompanied by ignore instructions. Capability inventory: Ability to create new pages and search existing notes. Sanitization: No evidence of input validation or output escaping for data retrieved from the application.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 04:20 PM