Skills
skills/tul-sh/skills/qwen-image/Gen Agent Trust Hub

qwen-image

Fail

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The Quick Start section contains the command curl -fsSL https://cli.inference.sh | sh, which downloads and executes a script from a remote server directly in the shell. This is an unverified remote code execution pattern.
  • [EXTERNAL_DOWNLOADS]: The skill downloads binaries and metadata from dist.inference.sh and cli.inference.sh during the setup and execution process.
  • [COMMAND_EXECUTION]: The skill requires the Bash tool to execute the infsh command-line utility for generating and editing images.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. 1. Ingestion points: The prompt and reference_images fields in the input for the infsh tool. 2. Boundary markers: Absent; user input is interpolated directly into a JSON string passed to the CLI. 3. Capability inventory: Permission to execute Bash(infsh *). 4. Sanitization: There is no evidence of sanitization or escaping of the user-provided prompt before it is passed to the command line, which could lead to command or instruction injection if the agent handles quoting incorrectly.
Recommendations
  • HIGH: Downloads and executes remote code from: https://cli.inference.sh - DO NOT USE without thorough review
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 5, 2026, 09:42 PM