ClawHub
Back to skill

Security audit

Git基础版

Security checks across malware telemetry and agentic risk

Overview

This is a straightforward Git helper skill whose repository-changing and remote Git actions are expected for its purpose, but users should confirm commands before execution.

Install only if you want an agent to help run Git commands in your repositories. Review any generated command before allowing commits, branch changes, pulls, merges, or pushes, especially when a remote repository or stored Git credentials may be used.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill explicitly instructs the agent to execute state-changing Git commands such as commit, checkout, merge, pull, and push, but it does not require explicit user confirmation or prominently warn that these actions modify repository state and may overwrite local context or publish changes. In an agent setting with exec capability, this creates a real risk of unintended repository modification, branch changes, merges, or remote publication caused by ambiguous prompts or misinterpreted intent.

Missing User Warnings

Medium
Confidence
83% confidence
Finding
The skill supports remote operations including clone, pull, fetch, and push and discusses credential errors, but it does not warn that these actions may expose repository URLs, branch names, commit metadata, or trigger use of stored SSH keys or tokens. In an agent context, remote Git actions can unintentionally transmit code or metadata to external systems or operate with sensitive credentials without explicit user awareness.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.