vercel-ai-sdk

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

• Third-party content exposure detected (high risk: 0.80). The skill's documentation and workflows show server-side tools that call web search/fetch functions (e.g., searchTool using searchWeb(query) and fetchWeather in references/tools.md and references/streaming.md) and message parts that accept external URLs (SourceUrlUIPart/FileUIPart in references/messages.md), which means the agent can ingest untrusted public web content that may influence tool calls and subsequent behavior.