Taizi Claw Shell

• Outcome: Review
• Latest audit: May 29, 2026 at 9:25 PM UTC
• Version:
  1.0.0

Overview

This skill openly provides a local shell tool, but its safety controls and tmux boundary are too weak for that level of access.

Install only if you intentionally want an agent to run local shell commands. Treat it as a raw shell interface, not a safe command runner; do not rely on its dangerous-command detection, and avoid using it around secrets or sensitive directories.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The skill executes arbitrary shell commands provided in input with only a weak denylist and no user-facing confirmation or disclosure for most commands. In an agent context, this is dangerous because benign-looking commands can still exfiltrate data, modify files, or run destructive shell features that bypass the simple keyword filter.

VirusTotal

62/62 vendors flagged this skill as clean.