The Agent Skills Directory

[SAFE]: No security findings were identified. The skill focuses on providing best practices for statistical experimentation, documentation templates, and growth program management.
[SAFE]: The skill was analyzed for 11 threat categories, including prompt injection, data exfiltration, and remote code execution. All external references are to well-known and reputable industry tools (Evan Miller, Optimizely, VWO) and do not involve automated script execution or package installation.
[SAFE]: Data ingestion is limited to local project context files (such as product-marketing.md) for information gathering. This standard practice does not represent a significant risk as the skill lacks executable capabilities, network operations, or file system modifications.

ab-testing