ClawHub
Back to skill

Security audit

Architecture Diagram Generator (FMTWiki Fork)

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent architecture-diagram generator with minor disclosure notes around saved HTML output and Google Fonts loading.

This skill is reasonable to install for generating architecture diagrams. Be aware that it saves generated diagrams as local HTML under /tmp, and opening the template may contact Google Fonts; review generated files before sharing because they can contain system architecture details you supplied.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Hidden Instructions

High
Category
Prompt Injection
Content
</defs>
        <rect width="100%" height="100%" fill="url(#grid)" />

        <!-- ─── COPY & CUSTOMIZE COMPONENTS BELOW ─── -->

        <!-- Region Boundary (dashed amber) -->
        <rect x="160" y="40" width="820" height="600" rx="12"
Confidence
70% confidence
Finding
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The skill description is broadly scoped to many kinds of technical diagrams and visualizations, which increases the chance of accidental over-activation on general user requests. In an agent setting, this can cause the model to invoke file-writing or deployment behavior when the user did not clearly request this specific skill, expanding the skill's operational footprint unnecessarily.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The 'When to Use' triggers are ambiguous and rely on broad phrases like 'visualize system components and relationships,' which can match many benign or unrelated requests. Without exclusion criteria, the agent may select this skill inappropriately, leading to unintended generation of standalone HTML artifacts and possible unnecessary tool usage.

Missing User Warnings

Low
Confidence
84% confidence
Finding
The skill instructs the agent to deploy output to a local /tmp path, which is a filesystem write action, but it does not warn about local file creation, overwrite behavior, or require user consent. While /tmp is typically low sensitivity, silent file output still enlarges the skill's side effects and could surprise users or conflict with host environment expectations.

Natural-Language Policy Violations

Low
Confidence
92% confidence
Finding
The document sets lang="en", which fixes the template to English without offering any language or locale choice. Under the stated policy, forcing a specific language without user opt-in is a natural-language locale constraint.

Missing User Warnings

Low
Confidence
81% confidence
Finding
This HTML template pulls a stylesheet from fonts.googleapis.com, which causes a network request to a third party when the page is loaded. The file contains no visible notice, comment, or other disclosure that loading the template will contact an external service.

Static analysis

No suspicious patterns detected.