The Agent Skills Directory

[SAFE]: The skill is designed to enhance operational security and reliability by providing a structured, evidence-based framework for deployment verification.
[COMMAND_EXECUTION]: The documentation suggests utilizing standard industry tools like kubectl, terraform, npm, pytest, and gh to perform safety checks. These commands are used for legitimate verification purposes and are contextualized within standard DevOps workflows.
[DATA_EXFILTRATION]: The skill promotes security best practices by explicitly recommending the sanitization of sensitive environment variables (e.g., using sed to mask values) before they are presented as evidence.
[PROMPT_INJECTION]: The skill is specifically engineered to resist behavioral overrides. It establishes "non-negotiable" safety mandates and provides a dedicated playbook to identify and counter common rationalization patterns used to bypass safety protocols.

verify-before-ship