The Agent Skills Directory

[SAFE]: The skill instructions focus exclusively on generating pedagogical questions based on code found in the conversation history. It does not define or request the use of any tools, network access, or file system modifications.
[SAFE]: The skill processes recently written code, which technically acts as an ingestion point for external data. However, the skill lacks any dangerous capabilities (such as shell execution or external data exfiltration) that could be leveraged by an indirect prompt injection attack, rendering the surface non-exploitable.
[SAFE]: The use of the $ARGUMENTS placeholder at the end of the file is a standard mechanism for passing user-provided topics to the agent. It is not used within any shell commands or dynamic execution contexts, posing no risk of command injection.

learn-quiz