ClawHub
Back to skill

Security audit

Code Dev Toolkit Free

Security checks across malware telemetry and agentic risk

Overview

This skill is a local coding-workflow guide with disclosed command and preference-memory behavior that appears user-directed and proportionate.

Install only if you are comfortable with a coding assistant that can read your current project, run local development commands when you ask, and keep explicit preferences in ~/code/memory.md. Do not ask it to remember secrets, credentials, private tokens, or sensitive project details.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
The skill claims it only provides guidance and does not automatically execute, yet it explicitly permits the exec tool and repeatedly instructs running shell commands and tests. This mismatch can mislead users and reviewers about the real execution capability, increasing the risk of unintended command execution in a code-development context where commands may modify files or run untrusted project code.

Intent-Code Divergence

Medium
Confidence
87% confidence
Finding
The privacy/security section says the tool does not automatically modify SKILL.md or auxiliary files, but elsewhere the skill describes creating and updating memory.md and other workspace support files. This inconsistency can cause users to underestimate what local files may be written, which matters because persistent memory files can store sensitive preferences or project-related data.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.