AGENT LAB: SKILLS

image-to-video

Fail

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION] (CRITICAL): The skill contains the command curl -fsSL https://cli.inference.sh | sh. This is a piped remote execution pattern that downloads and runs code from an untrusted source without any verification or integrity checks.
  • [EXTERNAL_DOWNLOADS] (MEDIUM): The skill uses npx skills add inference-sh/skills@... to download and install additional skills from an external repository. This introduces a supply-chain risk as the content of these external skills is not verified.
  • [COMMAND_EXECUTION] (MEDIUM): The skill requests broad permissions via allowed-tools: Bash(infsh *). This allows the agent to execute any sub-command and parameter for the infsh tool, which could be exploited if the tool itself has vulnerabilities or if the agent is manipulated into passing dangerous arguments.
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
CRITICAL
Analyzed
Feb 19, 2026, 03:41 AM