Monad Wordle Skill

• Outcome: Review
• Latest audit: May 28, 2026 at 1:18 PM UTC
• Version:
  1.0.0

Overview

This skill is a real blockchain game integration, but it gives an agent enough wallet authority to spend tokens on mainnet without sufficient built-in safeguards.

Install only if you intentionally want an agent to sign Monad Mainnet transactions for this game. Use a dedicated wallet with limited MON and WORDLE, verify the token, game contract, ABI, nad.fun flow, and API endpoint yourself, require manual approval for every buy, approval, and play transaction, and revoke or limit allowances after playing.

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
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Context-Inappropriate Capability

Medium

• Confidence: 91% confidence
• Finding: The skill advertises itself as a Wordle-playing HTTP API integration, but the documentation also instructs the agent to perform token trading on nad.fun. That expands the operational scope from gameplay into financial trading, increasing the risk of unintended asset purchases, slippage loss, and exposure to unreviewed third-party contract interactions.

Description-Behavior Mismatch

Medium

• Confidence: 93% confidence
• Finding: The manifest frames this as an HTTP game skill, but the execution flow requires wallet connection, token acquisition, ERC-20 approvals, and on-chain payment. This mismatch can mislead downstream systems or users into invoking the skill without understanding that it can trigger irreversible blockchain transactions and token spending.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill instructs agents to buy tokens, approve a contract, and pay to start a game without an explicit warning that these are irreversible financial actions. In agent settings, this is particularly dangerous because automated systems may execute approvals or purchases on behalf of a user without sufficiently informed consent.

Missing User Warnings

Low

• Confidence: 82% confidence
• Finding: The skill directs users to send wallet addresses and transaction hashes to a third-party API without any privacy or data-sharing notice. While these values are often public on-chain, centralizing them through an application endpoint can still enable tracking, profiling, and linkage of gameplay activity to wallet identities.

VirusTotal

64/64 vendors flagged this skill as clean.