The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute specific arkcli commands, such as arkcli models search, arkcli infer endpoint list --mine, and arkcli deploy. These operations are integral to the skill's purpose and are guarded by safety instructions, including the requirement for --dry-run flags and explicit user confirmation for any resource creation.
[SAFE]: The workflow design emphasizes security by requiring the agent to first check authentication and real-name status (via references to arkcli-auth) before proceeding with deployment. It also prevents unsupported scenarios, such as voice model onboarding, by implementing clear anti-trigger rules.

arkcli-onboard