review-contract
Warn
Audited by Snyk on Jun 14, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.85). The required workflow ingests the contract via “File upload”/“URL”/“Pasted text” (Step 1), and if the contract text comes from an outsider-authored source (e.g., the counterparty’s contract document or a URL to it), that readable prose is then loaded into the agent’s LLM context for clause-by-clause analysis (Step 4).
Issues (1)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata