aif-dockerize

Warn

Audited by Snyk on Mar 5, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.70). The skill's Step 4.1 explicitly requires using WebSearch to look up current stable versions of infrastructure images online ("verify current version of each infrastructure image" / "Search for ' docker official image latest version'"), which meaningfully influences generated compose.yml and therefore ingests untrusted, public third‑party content into the agent's decision flow.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 5, 2026, 12:13 AM