orca-cli
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPERSISTENCEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides numerous ways to execute shell commands and interact with terminal sessions via
orca terminal create,orca terminal send, andorca emulator exec.- [DATA_EXFILTRATION]: The skill includes anorca cookie getcommand which allows the agent to retrieve sensitive session cookies from the built-in browser context.- [PERSISTENCE]: The skill supports creating scheduled tasks usingorca automations create, allowing for long-term persistence of agent actions through daily, hourly, or cron-based triggers.- [DYNAMIC_EXECUTION]: Theorca evalcommand enables the execution of arbitrary JavaScript within the browser context, andorca execallows passing commands to internal agents.- [INDIRECT_PROMPT_INJECTION]: The skill's browsing capabilities (orca snapshot,orca goto) expose the agent to untrusted web content. The documentation explicitly acknowledges this risk, instructing the agent to treat page content as data and avoid executing commands derived from it.
Audit Metadata