ClawHub
Back to skill

Security audit

Smart Prompt Builder

Security checks across malware telemetry and agentic risk

Overview

This is mostly a normal prompt-building skill, but it bundles an LLM script that can send writing context and style data to DashScope without clear disclosure in the main docs.

Review before installing. Use scripts/build_prompt.py if you want local prompt construction. Do not run scripts/build_prompt_llm.py or set DASHSCOPE_API_KEY unless you intend to send your writing context and style profile to DashScope and accept any privacy, account, and cost implications; avoid passing sensitive or proprietary material unless that is acceptable.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
82% confidence
Finding
The skill documentation declares only local prompt-building behavior, yet the static analysis indicates capabilities for environment access, file read/write, and network use without any declared permissions. This is dangerous because users and orchestrators cannot accurately assess what data the skill may access or exfiltrate, especially when context, style profiles, or corpus data may be sensitive.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
This is a true security-relevant mismatch: the skill presents itself as a local prompt builder, but analysis shows it sends user context, style configuration, and possibly corpus contents to a third-party LLM endpoint using an API key from the environment. That creates a significant data disclosure risk because users may supply proprietary writing material or sensitive context without realizing it leaves the local environment.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.