youtube-thumbnail-design

Fail

Audited by Gen Agent Trust Hub on Feb 19, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • REMOTE_CODE_EXECUTION (CRITICAL): The skill instructs the user to execute curl -fsSL https://cli.inference.sh | sh in the 'Quick Start' section of SKILL.md. This pattern downloads and runs a script from the internet directly in the shell without any verification, integrity checks, or digital signatures. Since inference.sh is not in the list of trusted external sources, this is a critical security risk.
  • EXTERNAL_DOWNLOADS (HIGH): The skill relies on tools and scripts hosted at https://cli.inference.sh and https://inference.sh. Because these domains are untrusted, the skill is vulnerable to supply chain attacks where the remote content could be replaced with malicious code.
  • COMMAND_EXECUTION (MEDIUM): The skill requests permission for Bash(infsh *), which allows the agent to execute shell commands using the infsh utility. This provides a high-privilege capability that could be abused if the tool itself is malicious or if the agent is manipulated.
  • PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8) because it interpolates user-provided prompts directly into shell commands. 1. Ingestion points: Prompt values passed to infsh app run in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Arbitrary tool execution via Bash. 4. Sanitization: No sanitization or escaping is performed on the prompt input.
  • UNVERIFIABLE_DEPENDENCIES (MEDIUM): The skill documentation recommends adding further skills via npx skills add inferencesh/skills. These dependencies come from an untrusted registry and organization, increasing the attack surface.
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, 01:20 AM