ClawHub
Back to skill

Security audit

okx-dex-token

Security checks across malware telemetry and agentic risk

Overview

This skill is for OKX token data, but it under-declares command, network, streaming, auth, and possible payment behavior and uses overly broad activation terms.

Review before installing. This skill appears intended for token research rather than harmful activity, but it should be treated as a crypto data integration that may run local CLI commands, contact external services, stream WebSocket data, use environment-based authentication, and potentially trigger paid x402 flows. Install only if you trust the onchainos/OKX workflow and are comfortable with broad crypto-query activation terms being tightened or controlled by your agent.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
97% confidence
Finding
The manifest advertises only a read tool, but the body instructs the agent to run CLI commands and WebSocket operations. This capability mismatch can mislead policy enforcement, reviewers, or sandboxing logic into granting the skill more trust than warranted, enabling unintended command execution paths.

Intent-Code Divergence

Medium
Confidence
90% confidence
Finding
The documentation says no extra API key is needed, while elsewhere it references internal authentication via environment variables and possible paid x402 flows. Inconsistent auth and payment expectations can cause unsafe operator behavior, such as assuming anonymous operation, mishandling secrets, or being surprised by billable external requests.

Vague Triggers

High
Confidence
96% confidence
Finding
The trigger keywords are excessively broad, including generic terms like 'okx', 'level', 'token', 'data', 'skill', and common Chinese words. Overbroad activation can cause the skill to intercept unrelated conversations and steer the agent into reading local files or invoking external token-data workflows without sufficiently specific user intent.

Vague Triggers

High
Confidence
95% confidence
Finding
The repeated trigger section reinforces ambiguous activation terms without additional scope controls. Duplicating broad keywords increases the chance of accidental invocation and makes prompt-routing behavior harder to audit and predict.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.