The Agent Skills Directory

[Unverifiable Dependencies & Remote Code Execution] (MEDIUM): The skill instructs the user or agent to install software from non-standard sources.
It suggests npm i -g @kaban-board/cli. The @kaban-board organization is not in the trusted sources list.
It suggests brew install beshkenadze/tap/kaban, pointing to a personal GitHub repository tap which is also unverified.
[Indirect Prompt Injection] (LOW): The skill creates an attack surface by ingesting untrusted data from task titles and labels during kaban_status or kaban list operations. While it uses this data to suggest next steps (e.g., "Resume [task]?"), the current capabilities are limited to the Kanban workflow and mirroring to TodoWrite, minimizing the risk of high-severity side effects.

kaban-workflow