using-screenshots

Pass

Audited by Gen Agent Trust Hub on Oct 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • Command Execution: The skill utilizes the Bash tool to execute platform-specific screenshot utilities, including screencapture, gnome-screenshot, and scrot. While necessary for the skill's functionality, this involves the execution of system binaries via the shell.
  • Privilege Escalation: The instructions prompt the agent to suggest sudo commands (e.g., sudo apt install gnome-screenshot) to the user if required screenshot tools are missing. Although these are provided as suggestions for the user to execute, they involve administrative privileges.
  • Indirect Prompt Injection: The skill facilitates the ingestion of screen content, which is untrusted data that may contain malicious instructions designed to influence the agent's behavior.
  • Ingestion points: Screenshot files generated in /tmp/ and subsequently loaded into the agent's context using the Read tool as described in SKILL.md.
  • Boundary markers: The skill does not provide specific delimiters or instructions for the agent to disregard potentially malicious text found within the captured images.
  • Capability inventory: The agent uses the Bash tool for platform detection and screen capture, and the Read tool for visual analysis of the resulting files.
  • Sanitization: There is no mechanism described for sanitizing or validating text content identified within the screenshots before it is processed for analysis.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 12, 2026, 09:55 PM