Security audit

Axure Prototype Generator

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed Axure prototype generator that outputs active HTML/JavaScript examples, with no evidence of hidden network access, persistence, credential theft, or destructive behavior.

Install only if you are comfortable reviewing generated HTML/JavaScript before pasting it into Axure. Treat every generated prototype as active code, avoid entering real credentials or sensitive data into prototypes, and check for unexpected network requests or remote scripts if you modify or extend the generated output.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (9)

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The README invites extremely broad prompts such as generating dashboards or admin systems without defining safe scope, allowed behaviors, or forbidden content. In a skill that emits executable JavaScript/HTML for direct pasting into Axure, vague trigger boundaries increase the chance the model will produce unsafe active content, hidden external loads, or over-privileged prototype behaviors in response to untrusted user requests.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The README instructs users to paste generated JavaScript directly into Axure's JavaScript-backed inline frame flow, but provides no warning that this executes active code. Because the skill's core behavior is to generate executable HTML/JS, users may run unreviewed code that can open windows, load remote resources, exfiltrate data from the prototype context, or mislead users through active content.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The skill explicitly instructs users to paste and execute generated JavaScript via Axure's inline frame mechanism, which creates an active-code execution path with no safety guidance, sanitization boundaries, or trust restrictions. In this context, the skill is designed to generate arbitrary HTML/JS prototypes, so a malicious or compromised generation could execute script in the preview environment, load remote resources, or embed unsafe behaviors that users may run without review.

Description-Behavior Mismatch

Medium

Confidence: 97% confidence
Finding: The manifest says the skill generates Axure prototype code in JavaScript-format HTML for interactive prototype loading. This file does not generate prototype code from input or implement Axure-oriented generation logic; instead, it directly emits a specific stock portfolio dashboard with fixed business content and simulated refresh behavior.

Natural-Language Policy Violations

Medium

Confidence: 96% confidence
Finding: The HTML is explicitly emitted with lang='zh-CN' and the visible title/content is written only in Chinese, which imposes a specific language/locale on users. Under the policy, locale constraints should either be optional for the user or clearly justified as region-specific, neither of which is present here.

Natural-Language Policy Violations

Medium

Confidence: 90% confidence
Finding: This JavaScript generator produces user-facing titles, labels, and action text exclusively in Chinese, and the file comments also indicate a Chinese-only stock monitor page. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is clearly documented and justified, which is not present here.

Natural-Language Policy Violations

Low

Confidence: 85% confidence
Finding: The entire skill README is written as Chinese-only usage guidance and all example prompts assume Chinese interaction, with no indication that users may choose another language. This can be a natural-language policy issue when a skill implicitly enforces a specific language without offering user choice or documenting a justified locale restriction.

Natural-Language Policy Violations

Low

Confidence: 88% confidence
Finding: The phrase "中国化设计 - 红涨绿跌/中文排版/本地化组件" presents a specific locale and language convention as a default product behavior. Under the policy, forcing a specific language or locale without user opt-in is a natural-language policy issue unless clearly documented as region-specific or optional.

Natural-Language Policy Violations

Low

Confidence: 95% confidence
Finding: This JavaScript file hard-codes Chinese-language titles, labels, and placeholders such as '物流大厅', '线路查询', and '请输入'. Because the file provides no option to select another language and no justification that the skill is intentionally region-specific, it appears to impose a specific locale by default.

Static analysis

No suspicious patterns detected.