omni-admin
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs administrative tasks by executing commands via the
omniCLI, such asomni scim users-create,omni documents add-permits, andomni connections dbt-rotate-signing-key. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection because it ingests and processes external data to perform sensitive actions.
- Ingestion points: Data enters the context through command results from
omni scim users-list,omni schedules list, and user-provided CSV files viaomni uploads create. - Boundary markers: The instructions do not define clear delimiters or specific instructions for the agent to isolate external data from its operational logic.
- Capability inventory: The skill has significant capabilities, including managing user identities, adjusting access controls, and configuring database connections.
- Sanitization: There are no documented procedures for sanitizing or validating the content of the ingested data before the agent uses it for decision-making or further actions.
Audit Metadata