ClawHub
Back to skill

Security audit

Coding Cli Management

Security checks across malware telemetry and agentic risk

Overview

The skill has a coherent coding-delegation purpose, but it lets worker messages trigger powerful local AI coding tools in auto-approval mode.

Install only if you intentionally want trusted workers to delegate coding changes to the manager's authenticated local CLI tools. Use it in contained workspaces, restrict who can send coding requests, avoid secrets in prompts, and consider removing bypass flags or adding human review before running the CLI or overwriting shared workspace contents.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

High
Confidence
98% confidence
Finding
The script launches multiple AI coding CLIs with explicit safety-bypass flags such as --dangerously-skip-permissions, -y, and --yolo, allowing autonomous file modifications and other actions without interactive approval. In this context, the risk is amplified because the prompt file contents are fully injected into the agent and execution occurs inside a supplied workspace, so a malicious or compromised prompt can drive unchecked changes, destructive edits, or data exposure in that directory.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.