debugging

Warn

Audited by Snyk on Jul 8, 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 outsider-authored free text via the runtime-provided bug report artifacts (e.g., evals/files/*/report.md), which contain human-written descriptions/logs from parties other than the operating user.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 8, 2026, 08:11 PM
Issues
1
Security Audit — snyk — debugging