orca-emulator
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides logic for the agent to locate and execute the local
orcaCLI (or its variantsorca-devandorca-ide) to interact with mobile emulators. It includes specific environment-aware resolution logic to ensure the correct binary is used, particularly to avoid conflicts with system tools on Linux. - [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to run
ORCA skills get orca-emulatorto retrieve the complete set of operational instructions at runtime. This creates an attack surface where the agent processes external tool output as its primary instruction set. - Ingestion points: The output of the
ORCA skills get orca-emulatorcommand inSKILL.md. - Boundary markers: None explicitly defined in the discovery stub to separate fetched instructions from system prompts.
- Capability inventory: The CLI tool facilitates device booting, taps, gestures, hardware button control, camera injection, and accessibility tree access as described in
SKILL.md. - Sanitization: No specific sanitization or validation of the fetched instruction string is mentioned.
Audit Metadata