infinite-agentic-loop-claude

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill's installation explicitly clones https://github.com/disler/infinite-agentic-loop.git and the repository contains the .claude/commands/infinite.md command file that would be loaded and used at runtime to define/drive Claude slash-command prompts, so the remote repo is a runtime dependency that directly controls agent instructions.