The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute the command kamal config to verify the resolved deployment configuration. This is an expected and benign operation within the context of managing Kamal deployments.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill mentions reading sensitive local files such as ~/.ssh/config, ~/.ssh/id.pem, and .kamal/secrets. Accessing these files is required for the skill's primary function (SSH configuration), and there are no indicators of data being sent to external or unauthorized destinations.
[EXTERNAL_DOWNLOADS]: The skill includes links to official documentation for Kamal (kamal-deploy.org) and the SSHKit library on GitHub. These are well-known and reputable sources relevant to the skill's purpose.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from user-controlled configuration files like config/deploy.yml. While this constitutes a potential injection surface, the skill uses the data for standard configuration tasks and does not combine it with dangerous capabilities or provide paths for system-level exploitation.

ssh