data-visualization

Fail

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • Remote Code Execution (HIGH): A command was detected that downloads a script from a remote URL and pipes it directly into the shell for execution.
  • Evidence: curl -fsSL https://cli.inference.sh | sh.
  • Risk: This pattern allows a remote server to execute arbitrary commands with the privileges of the AI agent, potentially leading to full system compromise.
  • External Downloads (HIGH): The skill retrieves content from 'cli.inference.sh', which is not a verified trusted source.
  • Command Execution (HIGH): The script utilizes the system shell (sh) to process unvetted code retrieved over the network.
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
Feb 19, 2026, 09:06 PM