llama-analyst

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly ingests public third-party data as part of its runtime pipeline—e.g., "Data Sources (via data-orchestrator): DefiLlama API, Dune Analytics, Token Terminal"—and the agent is expected to read and interpret those external (including user-generated Dune) datasets for screening and ML predictions, so it can be exposed to untrusted third-party content.