audit-reply

Warn

Audited by Snyk on Jul 17, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.95). Outsider-authored free text from GitLab MR discussions (reviewer/commenter bodies) is fetched at runtime via glab api .../merge_requests/{iid}/discussions in scripts/fetch_mr.py, then returned as JSON and ingested into the agent’s LLM context for Step 3/4 analysis.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 17, 2026, 12:31 PM
Issues
1
Security Audit — snyk — audit-reply