The Agent Skills Directory

[COMMAND_EXECUTION]: Executes administrative commands to manage the agent environment, including restarting containers using docker stop and synchronizing skill files with rsync. These operations are directed at the platform's internal infrastructure.
[CREDENTIALS_UNSAFE]: Guides the setup of Vercel API credentials through a secure workflow. Tokens are stored in a dedicated secret manager (onecli) that uses proxy-based header injection, a robust practice that avoids leaking secrets in process lists or shell history.
[EXTERNAL_DOWNLOADS]: Configures the installation of the standard vercel package from the npm registry as part of the container build process.

add-vercel