cortex-code-tutorial

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The tutorial explicitly describes loading "remote skills" downloaded from URLs (e.g., GitHub) and shows the agent activating skills with $ (see Lesson 4 in SKILL.md and references/SKILLS_GUIDE.md), meaning the agent can fetch and execute third‑party skill instructions from public sources which can directly influence its behavior.