ClawHub
Back to skill

Security audit

Aliyun Qwen Deep Research

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed Alibaba Cloud research helper with expected local output and credential use, but it appears less complete than its description suggests.

Install only if you intend to use Alibaba Cloud Qwen Deep Research. Treat this as a request-preparation/support skill rather than a complete research executor, protect DASHSCOPE_API_KEY, use a virtual environment, and review or delete saved files under output/aliyun-qwen-deep-research/ when no longer needed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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 (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
77% confidence
Finding
The skill advertises and documents file-writing and network-oriented behavior, but it does not declare permissions or equivalent capability boundaries. In an agent ecosystem, undeclared capabilities reduce transparency and can cause the host or user to authorize a skill under false assumptions, increasing the chance of unintended data egress or persistent writes.

Tp4

High
Category
MCP Tool Poisoning
Confidence
89% confidence
Finding
The skill description claims a full deep-research workflow, but the documented behavior shown here appears limited to preparing a request payload and validation via compilation. This mismatch is dangerous because operators may trust the skill to have performed external research, evidence gathering, or reporting when it has not, which can lead to incorrect decisions and unsafe downstream automation.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill explicitly instructs saving research goals, confirmation answers, request payloads, and final report snapshots to disk, but it does not warn that these artifacts may contain sensitive user prompts, business data, credentials-adjacent context, or regulated information. In a research skill, this context makes the issue more dangerous because topics, attachments, and outputs are likely to include externally sourced and user-supplied sensitive material that can persist beyond the session.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.