Security audit

3d-generator

Security checks for vulnerabilities and agentic risk

Overview

This skill is a documented connector for external 3D model generation APIs, with privacy and redaction caveats but no hidden install, persistence, or destructive behavior found.

Install only if you are comfortable sending prompts, image URLs, and generated asset metadata to the OneKey gateway and downstream 3D providers. Keep the gateway access key private, avoid confidential or personal images unless your policy allows third-party processing, and prefer redacted or placeholder examples when sharing logs or documentation.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (3)

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The skill explicitly routes prompts and image URLs to external services/providers, but the documentation does not clearly warn users that their content may be transmitted to third parties and processed outside the local environment. This creates privacy and data-governance risk, especially if users submit proprietary images, personal data, or sensitive prompts under the assumption the agent handles them internally.

Ssd 3

Medium

Confidence: 88% confidence
Finding: The example output includes concrete session/workspace-style identifiers and provider-linked asset locations, which normalizes exposing traceable identifiers in public-facing documentation. Even if sample values are non-production, this pattern encourages leakage of user-correlated references that could aid enumeration, cross-tenant correlation, or accidental disclosure in downstream logs and tooling.

Ssd 3

High

Confidence: 97% confidence
Finding: The documented `raw_output` includes upstream object keys and signed original download links, which can expose internal storage layout and grant temporary direct access to provider-hosted artifacts. If such responses are surfaced to users or logs, attackers could retrieve assets, mine metadata, or replay signed URLs before expiration.

Static analysis

No suspicious patterns detected.