implement-review-loop

Warn

Audited by Snyk on Apr 3, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 1.00). The skill's workflow and scripts explicitly fetch and output GitHub PR comments (see Phase 5 in SKILL.md and scripts/fetch_pr_comments.sh / poll_pr_comments.sh which call gh api "repos/<owner/repo>/pulls/<pr>/comments" and return comment bodies for Claude to process), meaning untrusted, user-generated third‑party content is read and used to drive code changes and next actions.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 3, 2026, 11:02 AM
Issues
1
Security Audit — snyk — implement-review-loop