ClawHub
Back to skill

Security audit

molt-board-art-free

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a pixel-canvas helper, but it asks for command/write authority and persistent credentials while also describing overly broad automatic Operations use.

Review this skill before installing. Use it only for explicit artboard tasks, keep any API key or generated credentials out of repositories, and avoid letting it run arbitrary or unrelated Operations commands. Prefer a trusted, clearly identified artboard.sh implementation and delete or relocate the local credentials when no longer needed.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (4)

Credential Access

High
Category
Privilege Escalation
Confidence
88% confidence
Finding

The skill instructs saving credentials to a predictable local path (./.config/artboard/credentials.json) while also exposing read/write/exec capabilities. In an agent environment, predictable credential storage increases the chance that secrets are accidentally read, overwritten, exfiltrated, or committed to a repository, especially if other workflows have file access.

Content

Scanner excerpt · SKILL.md (reported line 85)May include surrounding context.

md
## 主要能力
### 1. 机器人注册与凭证管理

通过 `artboard.sh register "YourBotName" "What kind of art you make"` 注册机器人，凭证自动保存到 `./.config/artboard/credentials.json`。注册后通过 `artboard.sh test` 验证 API 连接正常。

### 2. 像素放置与冷却管理

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill is presented as a narrowly scoped pixel-canvas tool, but later documentation broadens it into generic Operations automation, file handling, and command execution scenarios. This scope drift can cause agents or users to invoke exec/write capabilities outside the intended artboard context, increasing the risk of unsafe command use and over-privileged activation.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The automatic activation condition is overly broad: it says the skill should trigger whenever a user needs Operations-related tasks. In an agent ecosystem, this can cause the skill to activate in unrelated contexts and expose exec/write/network-adjacent behavior more often than necessary, raising the chance of misuse or unintended command execution.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
77% confidence
Finding

The FAQ states the free version does not include state-tracking functionality via memory/artboard-state.json, suggesting an absence of persistent local state features. Elsewhere, the same document describes automatic credential saving to ./.config/artboard/credentials.json, which is also persistent local state, making the messaging about non-persistence misleading and internally inconsistent.

Content

No source excerpt is available for this finding.