braintrust-design-human-eval-review

Warn

Audited by Snyk on Aug 20, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (medium risk: 0.30). braintrust human review queues for per-criterion labels explicitly collect “free-text notes” and rationales at runtime, which can include outsider-authored text submitted into the review items that the agent then ingests to generate/update the golden set.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 20, 2026, 10:16 PM
Issues
1