ClawHub
Back to skill

Security audit

代理网络工具免费版

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent proxy tool, but it routes traffic through third-party proxy nodes with broad activation wording and limited privacy warnings.

Review before installing. Use this only when you explicitly want proxy generation or proxy-routed requests, avoid sending cookies, tokens, credentials, or proprietary data through the proxy, and treat both the TPN API key and generated proxy username/password as sensitive secrets.

SkillSpector

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

Intent-Code Divergence

Medium
Confidence
93% confidence
Finding
The skill states that secrets must not be echoed or logged, yet its example output includes live-style proxy credentials (username/password). Even if these are illustrative, normalizing credential display in responses encourages agents or users to expose ephemeral access tokens in chat logs, terminals, or telemetry, which can enable unauthorized proxy use until expiry and weaken overall secret-handling practices.

Vague Triggers

Medium
Confidence
82% confidence
Finding
The skill advertises extremely broad keywords like '代理', '工具', and generic networking phrases, creating weak activation boundaries. In agent environments, this can cause the skill to trigger in unrelated contexts and steer requests toward proxy generation or request forwarding, increasing the chance of unintended network operations or misuse for evasion-oriented tasks.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The coverage statement claims fragmented scenario keywords and broad semantic coverage rather than precise triggers, which makes the skill eligible for activation across ambiguous requests. Because this skill performs proxy-mediated network actions, ambiguous activation materially increases the risk of unintended third-party traffic routing, policy bypass attempts, or execution in contexts where the user did not clearly consent.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill facilitates credential generation and routes user traffic through third-party proxy nodes, yet it does not prominently warn about operator visibility, logging exposure, jurisdictional issues, or transmission of potentially sensitive request data. In this context, the omission is more dangerous because the tool is explicitly designed for geo-shifted and limit-evasion network access, where users may send authentication headers, cookies, or proprietary content through untrusted infrastructure.

Static analysis

No suspicious patterns detected.