The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a sentinel CLI tool to perform actions such as sync pull, scrub, sync push, and run finish. These are standard operational commands for the described workflow.
[DATA_EXFILTRATION]: While the skill's purpose is to share data externally, it implements several safeguards: it mandates a secret scrubbing process (sentinel scrub), requires a check for remaining secrets, and demands explicit user approval after showing exactly what will be published via git status.
[CREDENTIALS_UNSAFE]: The skill mentions the requirement of a token with workflow permissions for CI setup, but it does not attempt to harvest, hardcode, or store these credentials unsafely; it simply notes the necessity of the permission for the task.

sentinel-sync