ClawHub
Back to skill

Security audit

Miraix Meme Court

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly coherent, but it can automatically invoke an external crypto-trading workflow and send wallet or order details without a clear confirmation step.

Review this skill before installing. It does not appear malicious, but it can contact app.miraix.fun with wallet, trade, and order details. Use it only when you are comfortable sharing those fields with that service, and confirm the wallet, budget, risk mode, strategy, and any payload-preparation request before letting the agent run it.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill is designed to send a wallet address, budget, output contract, market, and potentially order identifiers to third-party APIs, but it does not instruct the agent to obtain explicit informed consent or warn the user about privacy and tracking implications. Because wallet addresses are persistent identifiers on a public blockchain, transmitting them to an external service can enable profiling of holdings and trading intent, and defaulting to a hard-coded wallet increases the risk of accidental disclosure or misuse.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
95% confidence
Finding

This call transmits user-linked trading inputs, including walletAddress and strategy parameters, to an external domain. In context, the transmission is core to the skill's functionality, but it is still security-relevant because the skill defaults missing inputs automatically and does not require an explicit permission step before sending potentially sensitive financial metadata off-platform.

Content

Scanner excerpt · SKILL.md (reported line 49)May include surrounding context.

  1. Run the desk:
bash
curl -sS https://app.miraix.fun/api/meme-rotation-desk \
  -H 'Content-Type: application/json' \
  -d '{
    "walletAddress":"<wallet-address>",

External Transmission

Medium
Category
Data Exfiltration
Confidence
97% confidence
Finding

The prepare endpoint sends walletAddress together with outputContract, market, amountUsd, mode, and symbol to generate an unsigned trade payload. Although unsigned payloads are safer than direct signing, this still exposes precise trading intent and could facilitate unwanted order preparation, surveillance, or misuse if triggered without clear user authorization.

Content

Scanner excerpt · SKILL.md (reported line 100)May include surrounding context.

then call:

bash
curl -sS https://app.miraix.fun/api/meme-rotation-desk/prepare \
  -H 'Content-Type: application/json' \
  -d '{
    "walletAddress":"<wallet-address>",

External Transmission

Medium
Category
Data Exfiltration
Confidence
91% confidence
Finding

The order-status request sends an orderId to a third-party API, which can disclose trading activity or allow correlation of user actions if intercepted or logged by the service. The exposed data is narrower than the desk or prepare calls, but it still involves external transmission of potentially sensitive transaction metadata without an explicit privacy warning.

Content

Scanner excerpt · SKILL.md (reported line 115)May include surrounding context.

  1. If the user gives an orderId and wants status, run:
bash
curl -sS https://app.miraix.fun/api/meme-rotation-desk/order-status \
  -H 'Content-Type: application/json' \
  -d '{
    "orderId":"<order-id>"

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill enables implicit invocation, but the activation scope in the metadata is broad and tied to trading-related tasks rather than a narrowly bounded trigger condition. This can cause the agent to invoke the skill in loosely related conversations and initiate safety checks, trade recommendations, or unsigned payload preparation without sufficiently explicit user intent, increasing the chance of unintended financial actions or sensitive workflow activation.

Content

No source excerpt is available for this finding.