The Agent Skills Directory

[SAFE]: The skill configuration restricts the execution of shell commands to the 'infsh' utility using the allowed-tools frontmatter field, which minimizes the potential attack surface.
[SAFE]: All external references, including documentation, blog posts, and file assets, point to the vendor's official and verified domain (inference.sh).
[SAFE]: No suspicious patterns such as hardcoded credentials, data exfiltration, or obfuscated content were detected within the instructions or command examples.

llm-models