ClawHub
Back to skill

Security audit

Openclaw

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed historical strategy game that uses expected local save files and optional LLM APIs, with only minor usability and privacy cautions.

Install only if you are comfortable with a game saving local campaign history under ~/.histrategy/rooms/ and, if enabled, sending turn text to the configured LLM provider. In shared chats, be aware that the visible delete command appears able to remove the shared saved campaign without a confirmation step.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Vague Triggers

Medium
Confidence
91% confidence
Finding
Triggers like 'rome' and 'triumvirate' are broad enough to activate during ordinary conversation unrelated to the skill. In an agent environment, unintended invocation can cause surprising behavior, unnecessary model/tool usage, or accidental state changes such as creating or modifying local game data.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The documentation explicitly states that game state is written to '~/.histrategy/rooms/' and survives restarts, but it does not pair that with a clear user warning or consent expectation. Silent local persistence can expose user-entered content, create privacy surprises, and leave residual data on shared or managed systems.

Missing User Warnings

Medium
Confidence
82% confidence
Finding
The delete command immediately removes the saved session for the entire chat with no confirmation, identity check, or ownership validation. In a group-chat multiplayer context, any participant who can invoke the skill may be able to erase the shared game state accidentally or maliciously, causing integrity loss and denial of service for the session.

Session Persistence

Medium
Category
Rogue Agent
Content
```python
from histrategy_sdk import Room

# Create an English Three Kingdoms game as Cao Cao
room = Room.create("my-campaign", faction="cao", lang="en")

# Get the intro scene
Confidence
84% confidence
Finding
Create an English Three Kingdoms game as Cao Cao room = Room.create("my-campaign", faction="cao", lang="en") # Get the intro scene intro = room.intro() print(intro["narrative"]) # Play a turn result

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
88% confidence
Finding
The trigger '三国' is extremely short and likely to appear in normal discussion of history, games, or literature. That raises the chance of accidental activation, which is less severe than code execution flaws but still risky because it can redirect the agent, invoke the skill unexpectedly, or cause unintended file-backed game actions.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.