review-fi-portfolio-workflow

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

• Third-party content exposure detected (medium risk: 0.65). At runtime the workflow calls external connectors (bond_price, yieldbook_bond_reference, yieldbook_cashflow, yieldbook_scenario, interest_rate_curve) that likely ingest and return vendor-provided reference text (e.g., security/issuer/ratings/call features) into the agent’s LLM context, which is outsider-authored content from third-party data providers.