churn-prevention
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains comprehensive instructional material for SaaS retention marketing. The content is aligned with its stated purpose and does not include prompt injection patterns or attempts to override agent safety guidelines.
- [COMMAND_EXECUTION]: The skill includes a local utility script,
scripts/churn_impact_calculator.py. Technical analysis of the script confirms it is a benign tool that uses only the Python standard library to perform mathematical revenue modeling. It does not perform network operations, execute shell commands, or access sensitive system files. - [SAFE]: The skill references established payment processors (Stripe, Braintree, Paddle) and third-party dunning tools for context, but does not request or handle API keys, credentials, or other sensitive authentication tokens.
- [SAFE]: The skill implements a context-aware design by reading an optional local file (
marketing-context.md). This ingestion is limited to providing business context for the agent's advice and does not pose a security risk through code execution or data exfiltration.
Audit Metadata