cleanup-properties

Pass

Audited by Gen Agent Trust Hub on Sep 29, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes custom property names and labels from a HubSpot instance, which serves as a potential surface for indirect prompt injection.
  • Ingestion points: Data is retrieved from the HubSpot CRM Properties API via GET requests as shown in SKILL.md.
  • Boundary markers: Instructions explicitly require the agent to obtain user confirmation before inventorying or performing any archive/delete operations.
  • Capability inventory: The skill utilizes network access to the HubSpot API and is capable of modifying or deleting CRM property schemas.
  • Sanitization: The provided logic excludes system-defined properties from the cleanup process.
  • [EXTERNAL_DOWNLOADS]: The documentation references the uv package manager hosted on the astral-sh GitHub repository, which is recognized as a well-known service for Python environment management.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 29, 2026, 05:21 PM