explainer-video-guide

Fail

Audited by Gen Agent Trust Hub on Feb 18, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • REMOTE_CODE_EXECUTION (CRITICAL): The skill instructs the user to run 'curl -fsSL https://cli.inference.sh | sh', which downloads and executes a script from an untrusted source without any integrity verification.
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill suggests installing additional unverified skills from 'inference-sh/skills' using 'npx', which presents a supply chain risk.
  • COMMAND_EXECUTION (MEDIUM): The skill provides bash command templates that use 'infsh' with user-provided prompts, creating a potential vector for command injection due to lack of sanitization.
  • PROMPT_INJECTION (LOW): The skill handles untrusted user input which could be used to manipulate agent behavior. 1. Ingestion points: User-supplied prompts and script content. 2. Boundary markers: No markers or protective instructions are used. 3. Capability inventory: Use of 'infsh' CLI tool via Bash. 4. Sanitization: No sanitization, validation, or escaping of inputs is mentioned.
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 18, 2026, 01:59 AM
Security Audit — agent-trust-hub — explainer-video-guide