blender-creation
Fail
Audited by Snyk on Nov 19, 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 legitimate asset-authoring pipeline, but its instruction to launch a background daemon with
BypassSandbox: trueweakens a sandbox protection beyond what the advertised preview needs, without malicious intent.
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (low risk: 0.10). The required workflow has the agent open an arbitrary user-supplied product URL in a browser subagent (and run
search_webfor text-only queries) and read the rendered page's titles, materials, and mechanism descriptions, so outsider-authored text is consumed only after the agent actively fetches it, which is the low tier.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (medium risk: 0.30). The only inbound fetch is
fetch_ambientcg_maps, which pulls CC0 texture zips from AmbientCG (a reputable third-party texture library, not a first-party publisher of this skill), so it is a common/reputable unofficial source; I'm applying recall-first to treat it as medium because the zip URL is chosen from a remote API response and the fetched PNGs are incorporated into the shipped asset.
MEDIUM W013: Attempt to modify system services in skill instructions.
- Attempt to modify system services in skill instructions detected (medium risk: 0.65). The skill is mainly a 3D asset-authoring pipeline, but it instructs the agent to run
sudo apt-get installfor Blender and to setBypassSandbox: trueto launch a background daemon, which touches sudo use and sandbox bypass.
Issues (4)
E004
CRITICALPrompt injection detected in skill instructions.
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
W012
MEDIUMUnverifiable external dependency detected (runtime URL that controls agent).
W013
MEDIUMAttempt to modify system services in skill instructions.
Audit Metadata