agents-sdk
Warn
Audited by Snyk on Feb 15, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.80). The skill explicitly instructs fetching live docs from a public GitHub URL (https://github.com/cloudflare/agents/tree/main/docs) and lists untrusted user-driven inputs such as "Email handling
- Receive and reply to emails" and streaming chat, meaning the agent may ingest and act on open/public third‑party or user‑generated content that could carry indirect prompt injections.
Audit Metadata