ai-sdk
Warn
Audited by Snyk on Feb 20, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.70). The SKILL.md "Otherwise" workflow explicitly tells the agent to use the public search endpoint (https://ai-sdk.dev/api/search-docs?q=your_query) and to fetch .md URLs such as https://ai-sdk.dev/docs/agents/building-agents.md, meaning the agent will fetch and read open/public third‑party web content which could influence its actions.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs fetching remote documentation at runtime from https://ai-sdk.dev (e.g., https://ai-sdk.dev/api/search-docs?q=your_query and linked .md like https://ai-sdk.dev/docs/agents/building-agents.md), which would load external text used to drive the agent's responses and is presented as the required source when local docs aren't available.
Audit Metadata