ClawHub
Back to skill

Security audit

君黎AI拆书

Security checks across malware telemetry and agentic risk

Overview

This skill appears to locally analyze a user-chosen novel text file and create local writing-reference outputs, with no evidence of hidden network access or destructive behavior.

Install only if you are comfortable letting the skill read the local text file you choose and store derived chunks, notes, outputs, and source-path metadata in a local project directory. Avoid using confidential or unpublished material unless that local persistence is acceptable, and review or delete the generated project files when finished.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Lp3

Medium
Category
MCP Least Privilege
Confidence
90% confidence
Finding
The skill instructs the agent to read a user-supplied local `.txt` file from an absolute path and to write multiple outputs under project directories, but the skill metadata does not declare those file access capabilities. This creates a permission-transparency gap: an operator may not realize the skill can access local files and persist data, increasing the risk of unintended exposure of sensitive local content or unsafe path usage.

VirusTotal

58/58 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.