ClawHub
Back to skill

Security audit

Git辅助

Security checks across malware telemetry and agentic risk

Overview

This Git helper is not overtly malicious, but it asks for command and write capability for repository-changing Git actions without clear user-confirmation boundaries.

Review this skill before installing. Use it only in repositories where you are comfortable with an agent running Git commands, and require confirmation for pull, push, branch changes, or any action that can alter local work or remotes.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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 (3)

Vague Triggers

Medium
Confidence
80% confidence
Finding
The description uses broad, mismatched trigger language such as system monitoring, log analysis, alerting, and deployment management for a Git helper. In agent environments, overly broad activation criteria can cause the skill to be invoked in unintended contexts, increasing the chance it executes repository-affecting commands when the user did not intend Git actions.

Vague Triggers

Medium
Confidence
81% confidence
Finding
The applicable-scope section is underspecified and broad enough that an agent may select this skill for generic ‘target data and configuration parameter’ tasks. Because the skill advertises exec capability and Git operations, ambiguous routing increases the risk of unintended command execution or repository modification in the wrong workflow.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill exposes the exec tool and describes Git pull/push/branch capabilities, but it does not prominently warn that it may execute shell commands and change repository state. In agent settings, missing consent and safety boundaries around state-changing actions can lead to unauthorized code changes, branch manipulation, pushes to remotes, or accidental disruption of working trees.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.