The Agent Skills Directory

[SAFE]: The skill performs legitimate file management tasks for maintaining evaluation baselines. Analysis confirms no unauthorized access to sensitive system files or environment variables. All operations are restricted to the 'promptops/runs/' and 'derived-index/baselines/' directories.
[SAFE]: No external dependencies, package installations, or remote code execution patterns were identified. The skill does not make network requests or download external content.
[SAFE]: No obfuscation techniques, such as Base64 encoding, zero-width characters, or homoglyphs, were found in the skill metadata or instructions.

apastra-baseline