ClawHub
Back to skill

Security audit

Claude Code Dispatch

Security checks across malware telemetry and agentic risk

Overview

This skill openly delegates coding work to Claude Code, but it gives a non-interactive external agent local file, shell, and inherited-secret access that users should review carefully.

Install only if you intentionally want OpenClaw to hand coding tasks to Claude Code with local repository access. Use the narrowest workdir, keep allowed tools minimal, avoid Bash unless needed, run with a clean environment that excludes unrelated secrets, and verify the Claude CLI installation and account before use.

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 (3)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill description and usage guidance promote dispatching tasks for file access, editing, and shell commands, but do not present a clear up-front warning that the delegated tool may modify host files and execute commands. In this context, omission is dangerous because the skill is specifically designed to cross the boundary from conversational assistance into direct host interaction.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The notes state that Claude Code inherits host environment variables, including API keys, but do not pair that with a strong privacy and secret-handling warning. Because this skill also enables shell execution, a delegated task could intentionally or accidentally expose, misuse, or exfiltrate sensitive credentials from the environment.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The script sends arbitrary prompt content directly to the external Claude CLI, which means potentially sensitive user data may be disclosed to a third-party model endpoint without an explicit warning or consent checkpoint at the point of execution. In this skill's context, that risk is heightened because the skill is specifically designed to dispatch coding tasks with file access and shell-related workflows, making it more likely that prompts include source code, secrets, proprietary paths, or operational details.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.