ClawHub
Back to skill

Security audit

Automation Scripts

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed automation-script guide, but it describes scheduled execution, automatic Git pushing, and automatic repair without enough scoping or approval safeguards.

Review before installing or using this skill for real automation. Only schedule scripts you understand, keep Git push behavior opt-in and scoped to approved repositories/remotes, require review before auto-repair actions, and verify logs/backups do not retain sensitive data longer than intended.

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
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Session Persistence

Medium
Category
Rogue Agent
Content
skill:automation-scripts --list

# 创建新脚本
skill:automation-scripts --create --name "my-script" --type "monitor"

# 执行脚本
skill:automation-scripts --run "script-name"
Confidence
60% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill explicitly describes a Git automation script that will automatically add, commit, and push changes, but it provides no warning, confirmation step, scope restriction, or discussion of the risk of transmitting local repository contents to a remote. In an automation skill, this is dangerous because users may unintentionally publish sensitive files, credentials, or incomplete work through unattended execution.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The OpenCode integration describes automatic diagnosis and repair after detecting service failures, but it does not warn that such remediation may change configuration, restart services, remove files, or otherwise alter system state. In the context of an automation skill, silent self-healing behavior increases the risk of destructive or cascading changes without operator review.

Natural-Language Policy Violations

Low
Confidence
92% confidence
Finding
The natural-language description is presented in Chinese and the file consistently assumes Chinese-language operation, with no indication that users may choose another language. Under the locale policy, forcing a specific language without opt-in or justification is a policy concern.

Static analysis

No suspicious patterns detected.