ms365-tenant-manager

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data from CSV files (e.g., new_users.csv, employees.csv) to perform bulk administrative actions.
  • Ingestion points: CSV data is read in SKILL.md, references/powershell-templates.md, and handled by the logic in scripts/user_management.py.
  • Boundary markers: The skill lacks explicit instructions to the agent to ignore potentially malicious instructions embedded within the CSV data fields.
  • Capability inventory: The skill possesses significant capabilities, including user creation (New-MgUser), license modification (Set-MgUserLicense), mailbox management (Set-Mailbox), and directory role management via the Microsoft Graph PowerShell SDK.
  • Sanitization: While scripts/user_management.py includes a validate_user_data function, it focuses on format validation (e.g., email structure, spaces in usernames) rather than content sanitization against prompt injection.
  • [DYNAMIC_EXECUTION]: The Python components (scripts/powershell_generator.py, scripts/tenant_setup.py, and scripts/user_management.py) utilize string templates and f-strings to dynamically assemble PowerShell scripts based on user-provided configuration and data. This generation of executable command strings is a core functional feature but represents a dynamic code construction pattern.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 02:33 PM
Security Audit — agent-trust-hub — ms365-tenant-manager