ClawHub
Back to skill

Security audit

hintme

Security checks for vulnerabilities and agentic risk

Overview

This skill is a straightforward game-hint helper with disclosed screenshot analysis and optional web lookup, and I did not find hidden commands, persistence, credential use, or destructive behavior.

Install this only if you want your agent to provide game hints from screenshots or descriptions. Be aware that it may use web search for game-guide context, and avoid sharing screenshots that include private chat, account names, or other personal information.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (6)

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · assets/banner.svg (reported line 21)May include surrounding context.

text
<rect width="1280" height="640" fill="url(#bg)"/>

  <!-- Ambient glow -->
  <circle cx="300" cy="320" r="250" fill="#00ff88" opacity="0.03"/>
  <circle cx="300" cy="320" r="150" fill="#00ff88" opacity="0.02"/>
  <circle cx="980" cy="320" r="250" fill="#b388ff" opacity="0.03"/>

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · assets/banner.svg (reported line 47)May include surrounding context.

text
<rect x="-55" y="0" width="110" height="36" rx="6" fill="#b388ff" opacity="0.1" stroke="#b388ff" stroke-width="1" stroke-opacity="0.2"/>
    <text x="0" y="24" font-family="system-ui, sans-serif" font-size="14" font-weight="600" fill="#b388ff" opacity="0.5">Hint 2: Clue</text>

    <!-- Hint 3 bar -->
    <rect x="70" y="0" width="130" height="36" rx="6" fill="#ffffff" opacity="0.05" stroke="#ffffff" stroke-width="1" stroke-opacity="0.1"/>
    <text x="135" y="24" font-family="system-ui, sans-serif" font-size="14" font-weight="600" fill="#ffffff" opacity="0.3">Hint 3: Answer</text>
  </g>

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · assets/banner.svg (reported line 55)May include surrounding context.

text
<!-- "hint" -->
  <text x="640" y="310" text-anchor="middle" font-family="system-ui, -apple-system, 'Segoe UI', Helvetica, Arial, sans-serif" font-size="140" font-weight="900" letter-spacing="-5" fill="url(#green)">hint</text>

  <!-- "me" -->
  <text x="640" y="420" text-anchor="middle" font-family="system-ui, -apple-system, 'Segoe UI', Helvetica, Arial, sans-serif" font-size="120" font-weight="300" letter-spacing="2" fill="#ffffff" opacity="0.8">me</text>

  <!-- Tagline -->

Hidden Instructions

High
Category
Prompt Injection
Confidence
70% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · assets/banner.svg (reported line 61)May include surrounding context.

text
<!-- Tagline -->
  <text x="640" y="490" text-anchor="middle" font-family="system-ui, -apple-system, 'Segoe UI', Helvetica, Arial, sans-serif" font-size="22" font-weight="400" fill="#ffffff" opacity="0.35" letter-spacing="4">STUCK IN A GAME? GET A HINT, NOT A SPOILER.</text>

  <!-- Bottom badges -->
  <g font-family="system-ui, sans-serif" font-size="13" fill="#ffffff" text-anchor="middle">
    <rect x="370" y="540" width="100" height="28" rx="5" fill="#ffffff" opacity="0.05" stroke="#ffffff" stroke-opacity="0.08" stroke-width="1"/>
    <text x="420" y="559" opacity="0.35">Any Game</text>

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest describes a skill centered on analyzing game screenshots and providing progressive hints, with no mention of external research or web access. README line L121 expands the behavior to include web search for game-specific knowledge, which is a broader operational capability than the stated screenshot-based hinting flow.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger phrases include very broad natural-language expressions like "I'm stuck" and "what do I do next," which can appear in many non-gaming conversations. This can cause the skill to activate in the wrong context, leading to unintended routing, irrelevant responses, or interference with other skills; while not directly a code-execution issue, it is a genuine security and safety concern because overbroad invocation reduces control over agent behavior.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.