The Agent Skills Directory

[SAFE]: The skill consists entirely of natural language instructions and structural templates for risk assessment. It does not perform any technical operations or include executable code.\n- [INDIRECT_PROMPT_INJECTION]: The skill uses the $ARGUMENTS placeholder to ingest user input. While this creates an entry point for untrusted data, the skill does not have access to any tools or system capabilities that could be exploited via prompt injection. There are no boundary markers for this input, but the lack of functional capabilities renders the risk negligible.

premortem