agents-sdk
Warn
Audited by Snyk on Aug 26, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The skill’s runtime path routes inbound HTTP/email/webhook/chat/RPC requests into agent handlers (e.g.,
routeAgentRequest,routeAgentEmail,onChatMessage,onRequest), so outsider-authored free text can be ingested into the LLM without any pre-selection step.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata