kde-rdp-desktop

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a local utility kli to manage a desktop-control daemon, capture screen content, and simulate user input. It also includes instructions for system diagnostics using systemctl, ss, and flatpak.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from the desktop framebuffer which could contain malicious instructions designed to manipulate the agent.
  • Ingestion points: Framebuffer captures from kli observe and kli observe region mentioned in SKILL.md and references/cli.md.
  • Boundary markers: None identified; there are no instructions to the agent to disregard instructions found within the visual data it processes.
  • Capability inventory: Subprocess calls to kli in references/cli.md allow for mouse movement, clicking, dragging, and typing, providing a significant surface for performing actions based on injected instructions.
  • Sanitization: No sanitization or validation of the screen content or CLI output is performed before the agent processes it.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 08:28 AM
Security Audit — agent-trust-hub — kde-rdp-desktop