orca-emulator
Fail
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: HIGHPRIVILEGE_ESCALATIONCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
- [PRIVILEGE_ESCALATION]: The skill provides explicit instructions to the AI agent to bypass sandbox restrictions. It states that if a command fails with 'runtime_access_denied', the agent should 're-run it with escalated permissions'.
- [COMMAND_EXECUTION]: The instructions require the agent to dynamically resolve and execute system binaries named 'orca', 'orca-dev', or 'orca-ide' depending on environment variables like 'ORCA_CLI_COMMAND' and the host operating system.
- [REMOTE_CODE_EXECUTION]: The skill serves as a stub that fetches additional instructions at runtime via the 'ORCA skills get orca-emulator' command. This represents dynamic instruction loading where the agent's behavior is determined by the output of an external command.
Recommendations
- AI detected serious security threats
Audit Metadata