ponder
Warn
Audited by Snyk on Mar 26, 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 configures Ponder to ingest public, user-generated on-chain events via remote RPC URLs (PONDER_RPC_URL_{chainId}) and exposes that data through the Ponder GraphQL API (queried via NEXT_PUBLIC_PONDER_URL), meaning the agent/app will read and act on arbitrary third‑party blockchain content as part of its workflow.
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata