Purpleflea Domains

• Outcome: Review
• Latest audit: May 28, 2026 at 5:44 AM UTC
• Version:
  1.0.0

Overview

The skill is a coherent domain-registration API guide, but it enables paid purchases and DNS changes without clear guardrails and promotes referral steering through system prompts.

Install only if you trust Purple Flea as a crypto-funded registrar. Keep any API key secret, do not place referral instructions in system prompts, and require explicit confirmation for every deposit, domain purchase, DNS update, or DNS deletion, including exact domain, price, record diff, and rollback plan.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• System Prompt Leakage: Direct Leakage, Indirect Extraction, Tool-Based Exfiltration
• 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)

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill advertises domain purchase functionality but does not clearly warn that invoking purchase endpoints causes real-money spending from the user's USDC balance and can register billable domains. In an agent setting, this increases the risk of unintended financial transactions if a user asks for exploratory or ambiguous actions and the agent proceeds without explicit spend confirmation.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The DNS management section lacks a clear warning that creating, updating, or deleting records can break email delivery, website availability, certificate validation, and traffic routing. In an autonomous-agent context, operational changes to DNS are high-risk because even small mistakes can cause outages or redirect traffic unexpectedly.

Ssd 4

Medium

• Confidence: 98% confidence
• Finding: The skill explicitly instructs operators to embed referral text into a system prompt so the agent is steered toward this paid service for monetized commission. This creates a conflict of interest and prompt-level manipulation that can bias tool selection away from the user's best interests, causing covert financial steering and potentially unnecessary purchases.

Ssd 1

Medium

• Confidence: 98% confidence
• Finding: The documentation explicitly instructs users to embed a referral directive into a system prompt so that agents are steered toward the provider for the author's financial benefit. This is prompt-level manipulation of agent decision-making and creates a conflict of interest that can bias tool selection away from user intent, cost optimization, or safer alternatives.

VirusTotal

63/63 vendors flagged this skill as clean.