The Agent Skills Directory

[SAFE]: The skill provides a structured diagnostic workflow for inference debugging using the 'openshell' CLI. All identified operations are consistent with the stated purpose of managing and verifying provider attachments and sandbox environments.
[SAFE]: Includes explicit security guidance to keep credentials and sensitive references out of diagnostic output, demonstrating alignment with safety best practices.
[SAFE]: No obfuscation, remote code execution, or unauthorized network activity was detected. Referenced URLs point to official documentation from the vendor.

debug-inference