The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted input from external documents and user text to generate execution plans. This represents an indirect prompt injection surface, though the skill's structured multi-phase process and self-review steps are designed to maintain control over the output.
[EXTERNAL_DOWNLOADS]: The skill references the official PlantUML public server for rendering and validating architecture diagrams. While this involves sending structural data to a third-party service, PlantUML is a well-known industry standard for this purpose.
[COMMAND_EXECUTION]: The skill generates instructions containing git and test commands intended for subsequent implementation phases. These commands are for documentation and handoff purposes and are not executed by the skill itself.

ss-plan