Miraix Binance Agent Firewall

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:21 PM UTC
• Version:
  1.0.0

Overview

This is a disclosed Binance prompt-audit skill whose main risk is that the trading prompt you provide is sent to Miraix's public API.

Install only if you are comfortable sending the trading prompt you provide to Miraix. Remove API keys, balances, personal information, proprietary strategy details, and private account data before use, and treat any permission plan as advisory rather than authorization to enable risky Binance permissions.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• 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
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill instructs the agent to send the user's raw trading prompt and symbols to a third-party public endpoint, but it does not require user consent, disclosure, or any data minimization before transmission. Prompts may contain sensitive trading strategy, account intentions, API-related details, or proprietary information, so silent exfiltration to an external service creates a real confidentiality and privacy risk.

Vague Triggers

Medium

• Confidence: 95% confidence
• Finding: The manifest sets `allow_implicit_invocation: true`, which lets the platform invoke this skill without an explicit user request while the trigger conditions are not narrowly defined in this file. Because this skill makes security decisions about Binance trading prompts and account permissions, accidental or over-broad invocation could cause untrusted trading content to be evaluated or rewritten in sensitive account-management contexts, increasing the chance of unsafe automation or misleading approval signals.

VirusTotal

55/55 vendors flagged this skill as clean.