simulator-screenshot
Warn
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION] (MEDIUM): The skill executes a local script "scripts/simulator-screenshot.sh" with arguments such as device IDs and file paths. If these arguments are derived from untrusted user input without proper escaping, it could lead to command injection.
- [EXTERNAL_DOWNLOADS] (LOW): Requires system-level tools like adb, jq, and Xcode Command Line Tools. While these are common developer utilities, they are external dependencies whose provenance is not managed by the skill.
- [INDIRECT_PROMPT_INJECTION] (MEDIUM): Evidence Chain: 1. Ingestion points: User-specified device names, IDs, and file paths. 2. Boundary markers: None specified in the instructions. 3. Capability inventory: Execution of local shell scripts via subprocess. 4. Sanitization: None. The lack of sanitization instructions for the agent when interpolating user-controlled strings into shell commands creates a risk of argument injection.
Audit Metadata