ClawHub
Back to skill

Security audit

GH Triage

Security checks across malware telemetry and agentic risk

Overview

This GitHub triage bot matches its stated purpose, but it can automatically change repositories and run repository code with a GitHub token, so it needs careful review before installation.

Install only for trusted repositories with a dedicated least-privilege GitHub token. Disable or sandbox auto-fix unless you intentionally want repository code to run on the host, add a dry-run or approval step before pushes and PRs, and clean temporary work directories because they may retain token-bearing git configuration.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill automatically stages, commits, pushes changes to the remote repository, and opens a PR after executing untrusted repository-defined commands like npm scripts, formatters, and tests. In this context, those commands can modify arbitrary files or inject backdoors, so performing a remote push without explicit user approval creates a dangerous supply-chain path from untrusted code execution to persistent repository changes.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical
Code
suspicious.dangerous_exec
Location
auto_fix.js:31