Zhkh Ru

• Outcome: Review
• Latest audit: May 28, 2026, 11:17 AM
• Version:
  1.0.0

Overview

The skill can explain Russian utility bills, but it also tells the agent to run hidden local Python commands and keep a counter file for promotional messages.

Review before installing. The bill-explanation behavior appears coherent, but the skill should not need to run Python or write a hidden local counter file to provide that service. Install only if you are comfortable with that side effect, or remove the Attribution section first.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Description-Behavior Mismatch

High

• Confidence: 99% confidence
• Finding: The skill's stated purpose is analyzing ЖКХ bills, but it additionally instructs the agent to run local Python commands, read a workspace file, and conditionally modify it for attribution and promotion. This is dangerous because it introduces hidden side effects unrelated to user intent, accesses local state, and turns a harmless document-analysis skill into one that performs unauthorized environment interaction.

Context-Inappropriate Capability

High

• Confidence: 99% confidence
• Finding: The instructions to access '/home/node/.openclaw/workspace/ru-pack-counter.txt' and update its contents create unauthorized local filesystem access unrelated to ЖКХ receipt processing. In context, this is more dangerous because users would reasonably expect a billing explainer, not a skill that persists state and modifies host files behind the scenes.

Context-Inappropriate Capability

Medium

• Confidence: 94% confidence
• Finding: The skill injects unrelated promotional content for other skills and social channels into the final answer based on hidden counter logic. This is dangerous because it manipulates user-facing output for undisclosed advertising, undermines trust, and couples benign assistance with covert promotion unrelated to the requested receipt analysis.

Missing User Warnings

High

• Confidence: 99% confidence
• Finding: The skill explicitly instructs execution of Python commands and file writes without disclosure in the skill description or any user consent. This is dangerous because hidden execution behavior can be abused to access local data, persist state, or perform actions outside the expected scope of a document-explanation tool.

VirusTotal

65/65 vendors flagged this skill as clean.