The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The guide includes instructions to use standard security tools and adds official repositories, such as the Falco Helm chart repository. These references target well-known and reputable organizations in the cloud-native security ecosystem.
[NO_CODE]: The skill consists entirely of Markdown documentation and configuration examples. It does not include any executable scripts or binary files, significantly reducing its direct attack surface.
[INDIRECT_PROMPT_INJECTION]: The skill acts as a security auditor, which inherently involves processing external data. Ingestion points: Analyzes user-provided configuration files (YAML, Dockerfiles) and runtime data (Falco logs). Boundary markers: The instructions lack explicit directives to treat external data as untrusted text rather than instructions. Capability inventory: The skill suggests using powerful tools like kubectl and helm to interact with the environment. Sanitization: No specific sanitization steps are defined for the data being analyzed.
[SAFE]: The skill provides defensive security guidance and implementation examples for container and Kubernetes environments. The instructions focus on hardening and reducing attack surfaces, such as dropping capabilities and enforcing non-root execution.

container-security