ClawHub
Back to skill

Security audit

cross-ref

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed GitHub repository analysis helper that reads PR and issue data and only posts comments after explicit approval.

Before installing, confirm you are comfortable letting the skill read PR and issue metadata from the named repository and send compact indexes to subagents. Use a read-only GitHub token for analysis, grant write access only for an approved posting run, and review every generated comment before creating the approval file and running with --execute.

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
88% confidence
Finding
The skill performs filesystem, network, and environment-backed operations but does not declare permissions, which weakens reviewability and policy enforcement. In this skill, those capabilities are expected for GitHub auditing, but the undeclared scope could let the agent access or write more than operators realize, especially because it invokes shell scripts, gh commands, and writes approval/workspace files.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.