ClawHub Skill 发布避坑指南

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:13 PM UTC
• Version:
  1.2.1

Overview

This is a Markdown-only publishing guide with disclosed manual examples and no hidden execution behavior.

Install this only if you want ClawHub skill publishing guidance. Review commands before letting an agent run them, especially examples that read your local ClawHub token, submit files, or set acceptLicenseTerms. The publisher should narrow the generic triggers to avoid accidental activation.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: Several triggers are extremely broad, including generic terms like `发布` and `更新`, which can cause the skill to activate for unrelated user requests. In an agent ecosystem, unintended activation can expose users to incorrect guidance, override more relevant skills, or create an avenue for prompt-surface abuse by matching common phrases too often.

VirusTotal

62/62 vendors flagged this skill as clean.