The Agent Skills Directory

[SAFE]: The skill consists of instructional content for the 'VERIFY' lifecycle stage. No malicious patterns such as prompt injection, obfuscation, or data exfiltration were found.
[NO_CODE]: There are no scripts, binaries, or executable components included in this skill. All content is descriptive markdown intended to guide the agent's reasoning process during quality assurance tasks.
[REMOTE_CODE_EXECUTION]: While the text mentions 'package install' in a procedural step, it is described as a 'structural mutation' to be verified rather than a command to be executed by the skill itself.

runtime-lifecycle