The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the management of cloud resources by executing the official Volcengine CLI (ve) and several Python and Bash helper scripts.
[SAFE]: Implements a secure multi-step OAuth login flow via scripts/ve_login_remote.sh, which uses a named pipe (FIFO) with restricted permissions (mkfifo -m 600) to feed authorization codes to a running subprocess without exposing them in command-line arguments.
[SAFE]: Enforces a strict read/write/destructive operation classification system, requiring explicit user confirmation before executing any command that creates, modifies, or deletes infrastructure.
[EXTERNAL_DOWNLOADS]: Fetches API schemas and documentation from official vendor-owned domains (api.volcengine.com) to assist in parameter discovery.
[CREDENTIALS_UNSAFE]: Contains explicit safety rules prohibiting the agent from reading the local configuration file (~/.volcengine/config.json) to prevent accidental exposure of long-term secrets.

volcengine-cli