鲁班 | Luban 打磨工坊

• Outcome: Pass
• Latest audit: Jun 11, 2026 at 7:52 PM UTC
• Version:
  2.0.0

Overview

This is a markdown-only workflow for improving and publishing skills, with disclosed research and editing behavior that users should keep under explicit control.

Install this if you want an opinionated assistant workflow for making a skill clearer, more testable, and easier to publish. Use it carefully on private repositories: ask for offline-only review when needed, confirm before network research, and require explicit approval before commits, pushes, releases, or deployments.

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: 89% confidence
• Finding: The trigger scope is very broad and matches generic requests like improving, optimizing, auditing, or publishing a skill. This can cause unintended invocation on adjacent tasks, leading the agent to perform extensive repository inspection, network research, and modification-oriented behavior when the user may have only wanted a narrow review.

VirusTotal

64/64 vendors flagged this skill as clean.