宠物冒险生活

• Outcome: Pass
• Latest audit: Aug 8, 2026 · 6:08 PM UTC
• Version: 1.0.0

Overview

This skill is a local pet diary simulator that writes disclosed game state files and optionally fetches weather, with no evidence of hidden data theft or destructive behavior.

Install only if you want a persistent local pet diary in your workspace. Expect it to create and update a pet-life/ folder and, unless you run advance with --offline, to contact Open-Meteo for weather for the pet's in-game location.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (6)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 89% confidence
• Finding: The skill documentation clearly instructs use of Python scripts that can read and write workspace files, invoke shell commands, and optionally access network-backed data such as weather, yet the manifest shown here declares no explicit permissions. That mismatch can cause the host or user to invoke a skill without understanding its real capabilities, increasing the risk of unintended file modification, command execution, or data exposure.

Vague Triggers

Medium

• Confidence: 83% confidence
• Finding: The description uses broad trigger language such as initializing, advancing, checking location, handling calls, and growing an emergent simulation, which could match ordinary conversational requests beyond the user's clear intent to run this specific skill. Over-broad invocation increases the chance that an agent auto-selects a file-writing, shell-using skill in situations where a simple textual answer would have been safer.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The skill states that the engine creates a `pet-life/` folder in the user's workspace and later documents multiple files it will write, but it does not prominently warn the user up front that invoking the skill causes persistent filesystem changes. In an agent environment, silent or weakly disclosed writes can surprise users, overwrite expectations for the workspace, and create privacy or integrity issues if sensitive directories are chosen.

Missing User Warnings

Low

• Confidence: 91% confidence
• Finding: The skill makes a live network request to a third-party weather API during normal operation without any user-facing notice or consent. While the transmitted data is limited to hardcoded latitude/longitude and timezone values, this still creates an undisclosed external dependency and leaks usage metadata such as request timing and selected in-game location.

Missing User Warnings

Low

• Confidence: 95% confidence
• Finding: The skill persistently writes state, world, and diary data into the workspace without explicit user disclosure at the point of use. This can surprise users by leaving durable files containing generated content and activity history, especially in shared or synced workspaces.

External Transmission

Medium

• Category: Data Exfiltration

Content

"timezone": location["timezone"],
        }
    )
    url = f"https://api.open-meteo.com/v1/forecast?{query}"
    try:
        with urllib.request.urlopen(url, timeout=8) as response:
            payload = json.loads(response.read().decode("utf-8"))

• Confidence: 94% confidence
• Finding: This code performs external transmission to api.open-meteo.com during diary advancement, creating a third-party data flow. Although it does not appear to send direct user secrets, it still exposes request metadata and ties skill usage to an external service, which is more sensitive because the skill otherwise operates on local state and may be assumed to be offline.

VirusTotal

65/65 vendors flagged this skill as clean.

Static analysis

No suspicious patterns detected.