The Agent Skills Directory

[SAFE]: No security issues detected. The skill provides legitimate guidance for model interpretability using the SHAP library without any malicious patterns or deceptive instructions.
[EXTERNAL_DOWNLOADS]: The skill references the shap package and its standard data science dependencies which are hosted on official package registries. Installation instructions use the uv package manager.
[PROMPT_INJECTION]: The skill does not contain instructions designed to override safety guidelines or agent behavior. While it involves processing user-provided model data, the instructions are focused on diagnostic visualization and do not involve high-risk capabilities such as network exfiltration or arbitrary system command execution.

shap