orca-emulator

Fail

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: HIGHPRIVILEGE_ESCALATIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [PRIVILEGE_ESCALATION]: The skill explicitly directs the agent to re-run commands with escalated permissions when encountering a 'runtime_access_denied' error, which is a direct instruction to bypass security sandboxes.
  • [COMMAND_EXECUTION]: Instructions guide the agent to resolve and execute system binaries (orca, orca-dev, orca-ide) to control simulator hardware, gestures, and device rotation.
  • [INDIRECT_PROMPT_INJECTION]: The skill uses ORCA skills get orca-emulator to load version-matched guides at runtime. This creates a vulnerability surface where unverified output from a CLI tool is ingested into the agent's context. Evidence Chain: (1) Ingestion point: ORCA skills get command output described in SKILL.md. (2) Boundary markers: None present. (3) Capability inventory: Subprocess execution via resolved ORCA binary and simulator control (taps, typing) described in SKILL.md. (4) Sanitization: No sanitization or validation of the dynamically fetched content is specified.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Oct 9, 2026, 02:32 AM