cover

Warn

Audited by Snyk on Sep 17, 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). The workflow executes test commands via Bash and monitors/reads the raw output and test failure text in Phase 4 and Phase 5 (heal-loop.js), which can contain outsider-authored free text from external dependencies or test failure messages.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 17, 2026, 02:14 AM
Issues
1
Security Audit — snyk — cover