ios-visual
Warn
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses shell access to run commands such as
xcrun simctlandnpx. It executes a specific TypeScript capture script located at a path resolved from theSCREENSHOTS_IOS_DIRenvironment variable. - [REMOTE_CODE_EXECUTION]: The skill executes local code via
npx tsx. The path to the source file is dynamically resolved at runtime using an environment variable, which is a form of dynamic code execution. - [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it analyzes screenshots and design images that may contain untrusted content. 1. Ingestion points: PNG screenshots from the simulator and user-provided design mockups (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: Full shell command execution via the
Bashtool. 4. Sanitization: No sanitization or validation of the visual content is performed before analysis.
Audit Metadata