churn-prevention
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill contains references to documentation and integration guides hosted on GitHub (e.g.,
github.com/coreyhaines31/marketingskills). These links are provided for informational purposes for the user or agent and do not involve automated remote code execution or script piping. - [INDIRECT_PROMPT_INJECTION]: The instructions prompt the agent to read local context files such as
.agents/product-marketing.mdif they exist. While this is an ingestion point for external data, the skill uses this information to personalize SaaS retention strategies. There are no patterns suggesting that this data is used in unsafe executable contexts or prompt-interpolated without boundaries. - [COMMAND_EXECUTION]: The skill mentions various CLI tools (e.g.,
stripe,customer-io,posthog) in a reference table for the user's awareness. It does not provide automated scripts or instructions that would cause the agent to execute these commands in a dangerous or unauthorized manner.
Audit Metadata