youtube-thumbnail-design
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 a shell command that pipes a remote script directly into the interpreter:
curl -fsSL https://cli.inference.sh | sh. This execution method is highly dangerous because the source script can be modified by the remote host to include malicious payloads without any local verification or review. Evidence found inSKILL.md. - External Downloads (MEDIUM): The skill uses
npx skills addto fetch and install additional functionality frominference-sh/skills. These remote resources are not from whitelisted trusted organizations and introduce unverified dependencies into the environment. - Command Execution (LOW): The skill configuration defines
allowed-tools: Bash(infsh *), which provides the agent with the authority to execute any subcommand for theinfshutility, expanding the attack surface if the utility itself has vulnerabilities.
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