churn-prevention

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill primarily consists of informational content and instructions for an AI to act as a SaaS retention expert. It does not include any scripts or executable code.
  • [SAFE]: Analysis of the instructions shows no evidence of prompt injection, obfuscation, or hardcoded credential exposure.
  • [SAFE]: The skill references standard development tools (e.g., Stripe CLI, PostHog) and other internal skills for coordination, which is expected behavior for a workflow-oriented agent skill.
  • [SAFE]: While the skill instructions specify reading local context files such as .agents/product-marketing.md, this is a constrained and documented behavior for providing contextual assistance within the agent's environment and does not involve external data exfiltration.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 12:30 AM
Security Audit — agent-trust-hub — churn-prevention