ohos-system-dev
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external device logs and UI layout information, which could contain malicious instructions designed to influence the agent's behavior.
- Ingestion points: Data is ingested from
devecocli log,devecocli ui layout, anddevecocli ui window list(SKILL.md). - Boundary markers: Absent; the skill does not provide specific instructions or delimiters to distinguish between log content and agent instructions.
- Capability inventory: The agent can execute build commands (
devecocli build) and system-level device controls (hdc shell reboot,param set ohos.startup.powerctrl reboot) (SKILL.md). - Sanitization: Absent; external log and UI data are processed without defined filtering or sanitization steps.
- [COMMAND_EXECUTION]: The skill invokes several command-line tools for environment verification, software building, and device management.
- Evidence: Uses
Get-Commandto verify tool locations,devecoclifor project builds and device inspection, andhdcorparamfor managing reboots and service persistence on the target hardware.
Audit Metadata