building-agents-with-eve

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

• Third-party content exposure detected (medium risk: 0.30). [Teaching mode] reads lesson text from the externally served Vercel Academy Lesson markdown bodies (fetched at runtime via GET https://vercel.com/academy/building-agents-with-eve/<lesson-slug>.md, including an <agent-instructions> block and full lesson body), so outsider-authored text can be ingested if the attacker can publish/affect that lesson content.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill's Teaching mode explicitly fetches lesson content at runtime from GET https://vercel.com/academy/building-agents-with-eve/.md (and related academy search endpoints) and instructs the agent to follow the returned block, so remote content directly controls the agent's instructions.