The Agent Skills Directory

[SAFE]: The skill provides instructional guidance for Docker best practices, including using multi-stage builds, non-root runtime users, and externalizing secrets. No evidence of prompt injection, data exfiltration, or malicious code execution was found.
[DATA_EXPOSURE_AND_EXFILTRATION]: The instructions explicitly forbid baking secrets into images, aligning with security best practices for credential management.
[INDIRECT_PROMPT_INJECTION]: While the skill reviews external code (Dockerfiles and Compose manifests), it does so to enforce safety boundaries and does not possess capabilities that could be exploited by malicious payloads in the processed data.

kernel-docker