The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the lightning CLI and the lightning-sdk Python library to create, start, stop, and manage cloud studios. This includes running shell commands and Python scripts on remote cloud machines.
[EXTERNAL_DOWNLOADS]: The skill manages its own environment by installing or upgrading the lightning-sdk package from the official Python package registry. These downloads are standard for the vendor's provided tools.
[INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes data from external sources, including cloud API responses and remote command outputs, creating a surface for indirect instructions.
Ingestion points: The agent reads command outputs via studio.run and logs using tail from remote studios, as well as metadata from lightning api calls.
Boundary markers: The instructions do not define specific delimiters or warnings to help the agent distinguish between data and potential instructions embedded within the logs or API responses.
Capability inventory: The agent possesses significant permissions, such as deleting studios (lightning studio delete), stopping compute resources, and transferring files (lightning cp).
Sanitization: No explicit logic is provided to sanitize or filter the retrieved log content or API data before the agent processes it.

lightning-studios