operate-vmos-edge-cli
Pass
Audited by Gen Agent Trust Hub on Apr 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill employs a CLI utility ('vmos-edge-cli') to control virtual devices and local application state. This includes functionality like 'device shell' for executing commands on Android systems and 'app start' for process management. A provided JavaScript helper ('scripts/ensure-installed.mjs') also executes 'npm' commands to manage the CLI's installation state.
- [EXTERNAL_DOWNLOADS]: The skill automates the installation of its primary CLI tool using the official NPM registry, a well-known service, ensuring that the necessary binaries are available to the agent.
- [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection due to its role in processing output from external environments. 1. Ingestion points: Untrusted data enters the agent context through 'ui state' (app content), 'device shell' (command output), and 'ui network' (network logs). 2. Boundary markers: The instructions do not specify delimiters or 'ignore' instructions for the data returned by the tools. 3. Capability inventory: The skill can execute device-level commands, interact with UI elements, and modify local configurations. 4. Sanitization: No explicit sanitization or filtering of external tool output is defined within the skill files.
Audit Metadata