ClawHub
Back to skill

Security audit

Soft Ip Full Lifecycle Zijian Clawhub Reviewfix

Security checks across malware telemetry and agentic risk

Overview

The skill appears purpose-aligned, but it under-discloses that the full user question is sent to a third-party service and saved locally in a sensitive legal/IP workflow.

Review this skill before installing. Avoid putting company secrets, source details, personal data, or legal document content in the initial question unless you accept that it will be sent to api.ideaidea.com.cn and saved in a local order file. Clear the local order files after use if the question or payment metadata is sensitive.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Tp4

High
Category
MCP Tool Poisoning
Confidence
88% confidence
Finding
The skill promises that only limited order metadata is stored locally, but its documented workflow includes passing the full user question into order creation and persisting outputs that include QUESTION, with the analyzer also indicating the full question is stored locally. That creates a transparency and data-minimization issue: users may provide commercially sensitive IP-registration details believing they are not being retained, leading to unexpected local exposure and possible disclosure through local compromise or shared workstations.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The script transmits the full user-provided question to a third-party remote API for order creation, and the code itself provides no explicit consent prompt, warning, or minimization before sending that content. Even though HTTPS is used and the skill metadata says no sensitive documents are uploaded, free-form questions can still contain personal, legal, or proprietary information, creating a privacy and data-governance risk if users are not clearly informed.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
This code sends a payment-related credential and order number to a third-party endpoint over the network, but the file contains no mechanism to inform the user at execution time that credential-bearing data will be transmitted. In a skill handling software IP assessment orders, silent transmission of billing or authorization material increases privacy and consent risk, especially because the service is third-party and the request is triggered automatically once local order data is loaded.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.