ClawHub
Back to skill

Security audit

clawtan

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed game-playing skill that uses an external Clawtan CLI, saves game session files, and keeps game notes, with no artifact-backed evidence of hidden or malicious behavior.

Install only if you trust the external npm clawtan package and are comfortable with the agent making game moves and sending public game chat. Treat saved Clawtan session files as game credentials, clear stale sessions when done, and review STRATEGY.md and HISTORY.md occasionally because the skill intentionally updates them after games.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill explicitly instructs the agent to append to `history.md` and rewrite `strategy.md`, which are local workspace files, without any warning or consent boundary around modifying user data. Even though these files are part of the skill, unconditional write instructions can cause unintended data loss or persistent workspace changes, especially because `rewrite` implies destructive replacement.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.