The Agent Skills Directory

[SAFE]: The skill is comprised entirely of markdown instructions and JSON metadata designed to assist an agent in generating handoff documentation. It does not define tools, scripts, or external dependencies.
[PROMPT_INJECTION]: The skill includes instructions to process external context such as work status and artifacts. While this creates a surface for indirect prompt injection, the risk is negligible as the skill lacks execution capabilities (tools or shell access).
[DATA_EXFILTRATION]: Instructions suggest identifying artifacts like logs and tickets for handoffs. This is a standard context-gathering practice for the stated purpose and does not involve automated exfiltration to external domains.

handoff-director