ClawHub
Back to skill

Security audit

Merge Check Tool Free

Security checks across malware telemetry and agentic risk

Overview

This skill is mostly a PR mergeability analyzer, but it asks to use authenticated GitHub access while also containing overbroad triggers and unrelated create/modify/delete capability language.

Review this carefully before installing. It appears intended for read-only PR analysis, but install/use should be limited to a specific PR you want analyzed. Avoid using it for generic data-analysis requests, and do not point it at private repositories unless you are comfortable with the agent querying PR metadata, files, comments, reviews, commits, checks, and repository metadata using your current gh login.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Confidence
76% confidence
Finding
The skill is positioned as a read-only PR analysis tool, yet its capability section claims support for create/modify/delete-style operations. In an agent environment with exec enabled, this ambiguity can broaden the model's perceived authorization and increase the chance that it performs unintended state-changing actions against local files or GitHub resources under the user's authenticated context.

Vague Triggers

High
Confidence
91% confidence
Finding
The trigger conditions are overly broad and describe generic data-analysis tasks rather than narrowly scoping this skill to PR mergeability assessment. In agent routing contexts, vague triggers can cause the skill to activate for unrelated requests, leading it to access GitHub-authenticated repository data unnecessarily through gh and expanding data exposure beyond user intent.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill does not prominently warn users that execution will query GitHub-authenticated data via the local gh CLI, potentially including private repository metadata accessible under the user's credentials. In an agent setting, missing disclosure weakens informed consent and can cause sensitive repository information to be fetched when the user may assume the operation is purely local or descriptive.

VirusTotal

55/55 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.