shottr
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill includes several scripts (Swift, Python, and Bash) designed to interact with macOS system utilities. These include
screencapturefor taking screenshots anddefaultsfor reading or writing application preferences. It also suggests usinglaunchctl asuserto handle GUI sessions in headless environments. - [INDIRECT_PROMPT_INJECTION]: The skill implements an OCR pipeline using Apple's Vision Framework (
VNRecognizeTextRequest) to extract text directly from the screen. This creates a vulnerability surface where an attacker could display malicious instructions on the screen to influence the agent's actions. - Ingestion points: Screen capture and OCR extraction logic in
screen_capture_ocr.swift. - Boundary markers: None; extracted text is printed directly to the console without delimiters or instructions for the agent to ignore embedded commands.
- Capability inventory: The skill provides scripts that can execute shell commands, write files to the Desktop, and modify application configuration strings.
- Sanitization: No filtering or sanitization of the OCR-extracted text is performed.
Audit Metadata