ai-sdk-model-manager

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

• Third-party content exposure detected (high risk: 0.90). The skill's Step 3 ("Research New Models") explicitly launches a "researcher subagent" that uses web search to fetch public provider documentation and web content, and that externally-sourced information is then used to set model parameters and decide which models to add, exposing the agent to untrusted third-party content.