Fbx To Glb Skill Repo

• Outcome: Review
• Latest audit: Jul 9, 2026 · 11:18 AM UTC
• Version: 1.0.1

Overview

This is a coherent FBX-to-GLB conversion skill, with ordinary local file and deployment behavior, but users should note its web template depends on third-party browser libraries.

Install this if you want a user-directed FBX-to-GLB converter and are comfortable with Node/npm for CLI use. For private or business-sensitive models, prefer the CLI or self-host a fully bundled web version; the provided web page loads Three.js from unpkg and the package appears to omit the documented assimpjs.wasm file, so the web template may need cleanup before production use.

SkillSpector

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration

Findings (5)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 90% confidence
• Finding: The skill instructs the agent to read arbitrary local FBX paths, run shell commands like npm install and node execution, and deploy web assets, yet it declares no permissions. This creates a capability/permission mismatch that can bypass user expectations and weakens safety controls around file access, command execution, and network-related deployment actions.

Context-Inappropriate Capability

Medium

• Confidence: 93% confidence
• Finding: The page claims conversion is purely local/no-upload, but it dynamically imports Three.js modules from unpkg at runtime. That creates a supply-chain and privacy boundary issue: visiting the page necessarily contacts a third-party CDN, and a compromised CDN response could execute arbitrary script in the page context. In a tool marketed for local-only processing of user files, this mismatch makes the issue more security-relevant because users may reasonably expect no external network dependency.

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The README suggests very broad natural-language activation phrases such as converting an FBX file or deploying a conversion website, which could plausibly match ordinary user conversation and cause unintended skill invocation. In an agent environment, overly generic triggers can route user data or file paths into this skill unexpectedly, increasing the chance of unintended processing or privilege use even without explicit malicious intent.

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The trigger phrase '3D 模型格式转换' is broad enough to match many unrelated model-conversion requests, which can cause unintended activation of a skill that reads files and suggests shell execution. In an agent environment, overbroad activation increases the chance of the wrong tool being invoked on sensitive local data or in inappropriate contexts.

Vague Triggers

Medium

• Confidence: 86% confidence
• Finding: The invocation description includes broad scenarios like generic batch 3D model conversion and website deployment, which can over-select this skill outside its validated scope. Because the skill entails local file handling, shell usage, and potential publishing steps, accidental activation meaningfully raises operational and security risk.

Static analysis

Detected: suspicious.dynamic_code_execution

Dynamic code execution detected.

Critical

• Code: suspicious.dynamic_code_execution
• Location: assets/web/assimpjs.js:8