ai-sdk

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

• Third-party content exposure detected (high risk: 0.90). The skill's instructions in SKILL.md and references/ai-gateway.md explicitly direct the agent to fetch and read public third-party content (e.g., searching ai-sdk.dev/docs and running curl against https://ai-gateway.vercel.sh/v1/models) which the agent must interpret to choose models and drive tool behavior, exposing it to untrusted public web content that can influence actions.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill explicitly instructs runtime fetches of external docs and model lists (e.g., fetching https://ai-sdk.dev/docs/agents/building-agents.md and curling https://ai-gateway.vercel.sh/v1/models) which are used to determine model IDs and authoritative behavior, meaning remote content can directly influence agent prompts/decisions.