The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the use of the ov CLI utility for resource management, filesystem navigation, and content modification.
[DATA_EXFILTRATION]: The ov backup and ov export commands allow for the aggregation and extraction of sensitive system state, user data, and session context from the viking:// namespace into portable .ovpack files.
[PROMPT_INJECTION]: The skill instructions allow the agent to ingest and process data from untrusted external sources, creating a risk for indirect prompt injection. 1. Ingestion points: ov add-resource accepts input from URLs, local files, and remote Git repositories. 2. Boundary markers: The skill does not provide instructions to isolate external content or treat it as untrusted data. 3. Capability inventory: The agent has access to powerful operations such as recursive deletion (ov rm --recursive) and data export. 4. Sanitization: There is no requirement for content validation or instruction filtering before the data is read by the agent.
[PERSISTENCE_MECHANISMS]: The ov task watch functionality allows for the creation of scheduled background tasks that periodically re-execute resource ingestion, effectively maintaining operations across agent sessions.
[EXTERNAL_DOWNLOADS]: The ov add-resource command enables downloading content from remote locations and cloning Git repositories, including those from the vendor's own namespace.
[PRIVILEGE_ESCALATION]: Documentation indicates that administrative operations such as backup, restore, export, and import require ROOT or ADMIN level permissions within the environment.

ov-resources