luxonis-device-setup

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill instructions include a reference to an official shell script installer (oakctl-installer.sh) from the vendor's release domain. The instructions explicitly direct the agent to ask the user to run the command themselves rather than executing it silently.
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the download of management tools and Python libraries (depthai) from official vendor-owned domains (luxonis.com subdomains) to enable hardware functionality.
  • [COMMAND_EXECUTION]: The agent performs local environment discovery and hardware diagnostics using Python imports and CLI tools (oakctl) to identify connected OAK devices.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface where it processes hardware and host-level facts during its discovery phase.
  • Ingestion points: System discovery data including OS details, hardware IDs, and network environment clues.
  • Boundary markers: Absent.
  • Capability inventory: Isolated package installation (pip), hardware CLI interaction, and local file creation (DEVICE.md).
  • Sanitization: Standard discovery output is used without specialized filtering for potential injection patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 01:17 AM
Security Audit — agent-trust-hub — luxonis-device-setup