cxxqt-rust-gui

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Vulnerability surface for indirect prompt injection in the command approval queue model. * Ingestion points: Command requests enter via the transport layer and are handled in the Rust backend state (Interactive-Terminal App Profile). * Boundary markers: No explicit sanitization or visual boundary markers for untrusted command metadata are demonstrated in the QML or bridge code snippets. * Capability inventory: Approved requests are routed for execution to terminal surfaces such as memory_terminal (referenced in Policy Cross-References). * Sanitization: The provided templates do not include logic for escaping or validating external content before display or processing.
  • [COMMAND_EXECUTION]: The skill outlines an architecture for an approval gateway that orchestrates the execution of commands on the host system via MCP.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 03:18 AM