Skills
▥AGENT LAB: SKILLS
skills/inference-sh-3/skills/video-ad-specs/Gen Agent Trust Hub

video-ad-specs

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's 'Quick Start' section includes the command curl -fsSL https://cli.inference.sh | sh. This is a highly insecure pattern that downloads a script from an untrusted remote server and executes it directly with shell privileges, allowing for full system compromise. \n- Unverifiable Dependencies (MEDIUM): The 'Related Skills' section uses npx skills add to download and install additional skills from inference-sh/skills. These external components are not from a trusted organization and could contain malicious code. \n- Command Execution (LOW): The skill is configured with allowed-tools: Bash(infsh *), which grants the agent permission to execute any subcommand of the infsh utility. While this is the intended use of the skill, it represents a significant capability that must be monitored. \n- Indirect Prompt Injection (LOW): The skill ingests untrusted user data into JSON payloads for the infsh CLI without explicit boundary markers or sanitization logic. \n
  • Ingestion points: The --input JSON strings in all infsh app run commands. \n
  • Boundary markers: None present. \n
  • Capability inventory: Execution of various remote AI apps via the infsh bash command. \n
  • Sanitization: None detected in the instructions.
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