ClawHub
Back to skill

Security audit

Qc Deep Feature Forensics

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed trading-analysis skill that reads an orders CSV, downloads market data, and writes local analysis files without evidence of hidden or unrelated behavior.

Run this only on trusted trading exports, preferably in a virtual environment and a dedicated working folder. Treat the generated CSV, markdown report, and cache as sensitive because they can reveal trading history, P&L patterns, and strategy signals.

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
88% confidence
Finding
The skill performs file-writing behavior (cache CSVs and generated reports) but does not declare any permissions for that capability. Undeclared write access weakens user visibility and policy enforcement, making it easier for a skill to persist data or overwrite files in ways the platform did not explicitly authorize. In this context the writes appear aligned with the skill’s function, but the lack of declaration is still a real security/control issue.

Tp4

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding
The declared description frames the skill as an analysis engine, but the documented behavior also includes network access, local persistence, trade reconstruction, ROI simulation, and report generation. This mismatch is security-relevant because users and policy systems may approve the skill expecting passive analysis while it actually fetches external data and writes files, increasing the attack surface for data exfiltration, privacy leakage, and unauthorized side effects. The skill context makes this more dangerous because it handles potentially sensitive trading history and creates persistent artifacts on disk.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.