The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands via olares-cli to perform file management tasks, including downloading content and deleting resources from the server's storage (drive/Home/...). These operations are protected by instructions requiring user confirmation.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources such as URLs, torrent metadata, and video titles, which creates a surface for potential injection attacks.
Ingestion points: External URLs, repository identifiers, and torrent files are provided as input to the create and inspect commands in references/olares-knowledge-download-lifecycle.md and references/olares-knowledge-download-torrent.md.
Boundary markers: The instructions advise the agent to quote URLs containing special characters, but there are no specific markers or warnings for handling the resulting metadata (e.g., titles, descriptions) returned by the CLI.
Capability inventory: The skill has the ability to create download tasks, remove files from the persistent volume, and modify server-wide settings.
Sanitization: No explicit sanitization or filtering of external metadata is described before it is displayed to the user or used in task naming.

olares-knowledge