The Agent Skills Directory

[SAFE]: The skill logic is confined to project management workflows and does not perform any high-risk operations. It does not execute shell commands, download external scripts, or modify system configurations.
[DATA_EXFILTRATION]: The skill reads project-related artifacts such as .agents/project-context.md and sprint boards to generate markdown summaries. This data remains within the agent's context and no network exfiltration patterns were identified.
[COMMAND_EXECUTION]: There are no patterns of arbitrary command execution or subprocess spawning. The workflow focuses on routing tasks to other project management subskills and providing status reports.

pm-execute