ClawHub
Back to skill

Security audit

Prediction Market Arbitrage Api

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly read-only arbitrage tooling, but it includes under-disclosed wallet and financial lookup commands beyond the advertised workflow.

Review before installing. The scripts appear read-only and do not execute trades, but they can query wallet-level financial data through AIsa if those commands are used. Use a limited AIsa API key, expect queries and wallet identifiers you provide to go to api.aisa.one, and only query wallet addresses you are authorized to analyze.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Tp4

High
Category
MCP Tool Poisoning
Confidence
93% confidence
Finding
The documented purpose is narrow—finding cross-platform arbitrage and checking liquidity—but the reported capabilities include access to wallet activity, orders/trade history, positions, wallet metrics, and P&L. That mismatch materially expands the data-access surface and could expose sensitive financial information unrelated to the user’s request, undermining informed consent and enabling overcollection.

Description-Behavior Mismatch

Medium
Confidence
89% confidence
Finding
The skill’s stated purpose is arbitrage discovery across markets, but this code also exposes Polymarket wallet activity, positions, wallet metadata, and P&L lookups for specific users. In an agent setting, this broadens access into user-level financial surveillance and can enable profiling or unauthorized querying of third-party wallet behavior beyond the declared scope.

Description-Behavior Mismatch

Medium
Confidence
84% confidence
Finding
The skill description emphasizes sports-market arbitrage, but the implementation exposes broad Polymarket and Kalshi market-query functionality without enforcing sports-only scope. This mismatch increases the chance that an agent can be repurposed for broader market intelligence collection than users or operators expect.

Missing User Warnings

Medium
Confidence
80% confidence
Finding
Wallet addresses and related identifiers are transmitted to a third-party API without any visible disclosure, consent flow, or minimization controls in the client. In a hosted agent environment, this can expose sensitive user-linked financial activity or enable third-party lookups on arbitrary wallets without clear user awareness.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.