skill-file-security

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

• Potentially malicious external URL detected (high risk: 0.90). The README shows runtime use of npx @netxeo/security-skill (https://www.npmjs.com/package/@netxeo/security-skill), which fetches and executes remote package code that installs AI instruction files and assistant configs (.skills/*.md, CLAUDE.md, .cursorrules, etc.) that directly control agent prompts/instructions.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)