The Agent Skills Directory

[SAFE]: The skill is primarily composed of domain-specific documentation, audit templates, and reference guides. It does not contain executable scripts or binary files.
[SAFE]: Section §0 of SKILL.md implements a mandatory safety gate that enforces HIPAA compliance by requiring confirmation of de-identification and synthetic data usage. It also includes explicit disclaimers regarding the limitation of its outputs to ML engineering review rather than clinical or actuarial certification.
[EXTERNAL_DOWNLOADS]: The README.md file contains an installation command for the 'skills.sh' platform. This is a well-known service for AI agent extensions, and the referenced resource belongs to the skill's author ('Yar177'), representing a standard vendor-owned installation path.
[PROMPT_INJECTION]: The skill is designed to audit user-provided ML notebooks and feature specifications (SKILL.md §1). While these represent an ingestion point for potentially untrusted data, the skill is restricted to generating markdown-based reports and YAML templates. It lacks the capabilities to execute commands or perform network operations, which significantly mitigates the risk of indirect prompt injection attacks.

claims-ml