ClawHub
Back to skill

Security audit

discord组合包

Security checks across malware telemetry and agentic risk

Overview

This communication bundle is broadly coherent, but it requests wide read, write, and execution-style authority across several messaging services without clear limits or safeguards.

Review this carefully before installing. It is not showing hidden malware or destructive code, but it can interact with several communication services and may read or send messages using account credentials. Install only if you understand which accounts it will access, restrict API keys to the minimum permissions needed, and avoid sending sensitive chats, email, attachments, or tokens through generic examples or logs.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
77% confidence
Finding
The documentation normalizes command execution and shell-style troubleshooting in a bundle whose stated purpose is communication. In a skill that already advertises exec capability, vague guidance to execute commands with user-supplied parameters can encourage unsafe operational use, increasing the risk of command injection or destructive local actions when operators or downstream agents follow the examples blindly.

Missing User Warnings

Medium
Confidence
83% confidence
Finding
The skill advertises read/exec/write capabilities across multiple communication tools without clearly surfacing the security consequences of transmitting data externally or running system-affecting actions. In this context, understated warnings can lead users or autonomous agents to pass sensitive content, invoke powerful actions, or mis-handle API credentials without understanding the trust boundary.

External Transmission

Medium
Category
Data Exfiltration
Content
api_url = "http://localhost:8080/api/skills/plug-bundle-discord"
headers = {"Authorization": "Bearer YOUR_API_KEY"}
# 执行技能
response = requests.post(api_url, json={
    "input": "示例输入数据",
    "options": {"mode": "advanced"}
})
Confidence
85% confidence
Finding
The example shows posting input data and an authorization bearer token to an HTTP API endpoint, which represents external transmission of potentially sensitive content. Even though the endpoint is localhost in the snippet, the pattern normalizes network transfer of user data and credentials without emphasizing TLS, secret protection, logging controls, or data-classification limits; in real deployments this can lead to credential exposure or unintended data exfiltration.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.