vercel-ai-sdk

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly defines a "search" tool that "Search the web for information" (Tools / Function Calling) and shows multi-step usage where web/search results are ingested and used in generateText, so the agent would fetch and act on untrusted public web content (search results) that could contain instructions influencing subsequent actions.