openai-image-vision

Warn

Audited by Snyk on Mar 9, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.80). The skill explicitly accepts HTTP(S) image URLs (see scripts/vision.sh URL branch and SKILL.md/README examples) and passes those arbitrary external images directly to the model for interpretation, meaning untrusted/user-hosted image content (including embedded text/instructions) can influence the agent's outputs and downstream actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 9, 2026, 04:18 AM