ClawHub
Back to skill

Security audit

GitHub开发规范工具

Security checks for vulnerabilities and agentic risk

Overview

This skill is not malicious, but it asks for broad development powers with vague scope and limited user-control guidance.

Review this skill before installing if you do not want a general development assistant to run commands or modify repository files. It appears aimed at legitimate GitHub workflow assistance, but users should limit it to trusted repositories, avoid exposing broad API keys, and require explicit approval before command execution, deployment, or file-changing actions.

SkillSpector

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
90% confidence
Finding
The skill advertises a generic `exec` capability while only documenting a conceptual whitelist restriction, with no concrete enforcement mechanism in the skill definition. This creates a dangerous mismatch: users or downstream agents may trust the safety claim and invoke shell execution in contexts where user-controlled input, repository content, or generated commands can lead to arbitrary command execution or destructive local actions.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The activation language is broad enough to match routine development, debugging, deployment, and code-generation requests, making the skill likely to trigger in many high-privilege situations. Because the skill also exposes read/write/exec capabilities, overly broad invocation increases the chance that an agent applies this powerful skill when the user did not intend shell execution or repository modifications.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill describes command execution, file handling, and API-key-based operations, but does not present prominent user-facing warnings about possible system changes, network effects, credential use, or destructive outcomes. In a skill with `exec`, `write`, and API interaction, missing impact disclosure can lead users and orchestrators to authorize risky actions without informed consent.

Static analysis

No suspicious patterns detected.