Lab — Adopt a Lab. Dog. 拉布拉多。Labrador.

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:18 PM UTC
• Version:
  1.0.3

Overview

This is an instruction-only virtual pet skill that uses disclosed AnimalHouse API calls, with no hidden local code or privileged system access.

Install only if you are comfortable using animalhouse.ai. Use non-sensitive profile, prompt, and care-note text, keep the bearer token private, and treat the release endpoint as potentially destructive even though the skill does not explain its exact effect.

SkillSpector

By NVIDIA

Vulnerability Patterns

• 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
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

• Confidence: 85% confidence
• Finding: The skill documents a destructive `DELETE /api/house/release` endpoint without warning the user that it is irreversible or advising any confirmation step. In an agent setting, this increases the chance of accidental data or asset loss if an automation, LLM, or user invokes the endpoint based only on endpoint availability.

VirusTotal

63/63 vendors flagged this skill as clean.