visualize-in-3d

Fail

Audited by Snyk on Nov 15, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E004: Prompt injection detected in skill instructions.

  • Potential prompt injection detected (medium risk: 0.30). The skill is a 3D-visualization workflow whose only notable concern is self-serving promotion of the author's own CodePawl datasets and Crucible npm/MCP install, which steers asset sourcing and installs toward the vendor without malicious intent, while downloads and executables are gated on user approval.

CRITICAL E005: Suspicious download URL detected in skill instructions.

  • Suspicious download URL detected (medium risk: 0.30). The skill directs the agent to fetch datasets from a single unverified publisher (codepawl on Hugging Face and GitHub), and to prefer the newest uploads with "the live card wins over this table," which bypasses the vetted list, but the visible content shows no concealed download-and-execute chain or decoy source, so the evidence is suspicious but not conclusive.

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

  • Third-party content exposure detected (low risk: 0.10). The required workflow actively searches and fetches outsider-authored web pages (news, secondary sources, Sketchfab/Poly Pizza model pages, and reference images) for research and asset sourcing, but no passively monitored inbox, feed, or ticket queue appears in the visible content, so the agent must select the items itself.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

  • Potentially malicious external URL detected (medium risk: 0.30). The decisive boundary is the setup step npx skills add codepawl/crucible in references/67js.md, which installs skill instructions from a GitHub repository, and publisher provenance is an unofficial third-party project (codepawl) with no enforced signature or digest, so it is a common/reputable unofficial dependency (medium); the Hugging Face GLB packs are loaded as data and do not trigger the gate.

Issues (4)

E004
CRITICAL

Prompt injection detected in skill instructions.

E005
CRITICAL

Suspicious download URL detected in skill instructions.

W011
MEDIUM

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

W012
MEDIUM

Unverifiable external dependency detected (runtime URL that controls agent).

Audit Metadata
Risk Level
CRITICAL
Analyzed
Nov 15, 2026, 10:00 PM
Issues
4