cleanup-workflows

Pass

Audited by Gen Agent Trust Hub on Apr 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized access attempts were detected in the skill instructions or code snippets.
  • [DATA_EXPOSURE]: The skill correctly uses environment variables (os.getenv('HUBSPOT_API_TOKEN')) to handle sensitive credentials, which is a standard and secure practice for API integration.
  • [EXTERNAL_DOWNLOADS]: The skill references the official HubSpot API domain (hubapi.com) for data retrieval, which is a well-known and trusted service for this context.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes workflow metadata (such as workflow names and states) retrieved from the HubSpot API. This constitutes a potential indirect prompt injection surface if the agent processes maliciously named workflows. However, the skill lacks dangerous capabilities that would make this exploitable.
  • Ingestion points: Workflow list retrieved from the HubSpot Automation API in SKILL.md.
  • Boundary markers: Absent.
  • Capability inventory: Includes HTTP GET requests via the requests library; no subprocess execution, file-writing, or dynamic code evaluation detected.
  • Sanitization: None explicitly defined for the external data returned by the API.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 9, 2026, 04:21 AM