legal

Warn

Audited by Snyk on Jul 23, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.75). Outsider free text can enter the LLM context via the runtime intake path “Accept the contract as file, pasted text, or URL reference” (i.e., user-provided contract content can be authored by the counterparty/outsider and then read by the model for analysis/redlines).

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 23, 2026, 09:18 PM
Issues
1
Security Audit — snyk — legal