ClawHub

Code Assistant

Security checks across malware telemetry and agentic risk

Overview

This coding assistant reads user-chosen code paths and advertises code-editing features, but the artifacts show no hidden execution, data transfer, or persistence.

Install only if you are comfortable letting it read the files or directories you ask it to analyze. Keep automatic fixes disabled unless you deliberately want edits, review diffs before accepting refactors or self-repair, and avoid pointing it at unrelated private directories or files containing secrets.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill prominently advertises refactoring, documentation, test generation, and optimization actions that can alter repository contents, but it does not clearly warn users that files may be created or modified. In a coding-assistant context, users may invoke these commands expecting advice only, leading to unintended source changes, broken builds, or accidental overwrites.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
Describing 'self-repair' without an explicit disclosure that it can automatically change code is risky because it normalizes autonomous modification of project files. In a tool integrated with coding agents, that increases the chance of silent or surprising edits that introduce regressions, security issues, or destructive changes in sensitive repositories.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal