MagicHaqi Pet Master

• Outcome: Review
• Latest audit: Jun 12, 2026 at 8:50 AM UTC
• Version:
  1.0.0

Overview

This skill is for a virtual pet game, but it asks the agent to handle your KeepWork password and token directly, so it should be reviewed carefully before use.

Install only if you are comfortable letting the agent handle your KeepWork login and game token. Prefer completing login yourself in the browser when possible, verify the MagicHaqi host before any token is used, do not reuse a high-value password, and explicitly approve adoption or any in-game spending.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Context-Inappropriate Capability

Medium

• Confidence: 88% confidence
• Finding: The skill expands from browser-only game operation into reading local workspace files containing pet persona or memory data. That creates an unnecessary filesystem access path not required for core gameplay, and if the environment exposes broader workspace contents, the agent could read unrelated sensitive data under the guise of role-play context.

Vague Triggers

Medium

• Confidence: 76% confidence
• Finding: The invocation language is broad enough to match ordinary pet-care or pet-chat requests, even when the user does not mean the MagicHaqi game. That can cause the skill to activate in the wrong context and then prompt for credentials, open external sites, or take actions on a game account unexpectedly.

Missing User Warnings

High

• Confidence: 95% confidence
• Finding: The skill instructs the agent to collect the user's username and password and send them to an external login API, but it does not require a clear consent/privacy warning or a safer delegated login flow. Even if the endpoint is legitimate, routing raw credentials through the agent materially increases credential-handling risk, phishing surface, and accidental secret exposure.

Missing User Warnings

Medium

• Confidence: 89% confidence
• Finding: The skill instructs the agent to submit a username and password to an external login API but provides no explicit guidance on credential minimization, secure handling, storage avoidance, or user warning/consent around transmitting account credentials. In an agent setting, this increases the risk of credential exposure, logging, replay, or misuse, especially because the agent is expected to operationalize the login flow on the user's behalf.

VirusTotal

64/64 vendors flagged this skill as clean.