ClawHub
Back to skill

Security audit

Binance Alpha Explorer

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Binance listing monitor that uses public Binance APIs and small local state files, with no evidence of credential theft, trading actions, or hidden exfiltration.

Install only if you are comfortable with the skill connecting to Binance public endpoints and keeping its known-symbol and alert history files under ~/.config/alpha. It should not need exchange credentials or wallet access; consider adding explicit permission metadata or containment if your environment requires strict network and filesystem disclosure.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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 (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
91% confidence
Finding
The skill documentation describes network access and persistent state in ~/.config/alpha, but no permissions are declared. This creates a transparency and trust problem: users or platforms may approve the skill expecting low-risk behavior while it can read/write local files and make outbound network connections. In this context the behavior appears aligned with the stated monitoring use case, so the issue is under-declaration rather than clearly malicious functionality.

Tp4

High
Category
MCP Tool Poisoning
Confidence
87% confidence
Finding
The skill metadata says it maintains the known symbols set in memory, but the file also documents persistence to disk, alert history storage, reset/status/history commands, and REST API use beyond the WebSocket-only description. This mismatch weakens informed consent and can conceal the real operational footprint, especially local data retention and additional network activity. The extra behaviors are not inherently dangerous in a Binance-monitoring skill, but inaccurate description increases risk because users may not expect stored data or broader capabilities.

Description-Behavior Mismatch

Medium
Confidence
89% confidence
Finding
The skill writes known symbols and alert history to local files under ~/.config/alpha even though the description says the detector maintains state in memory. This creates an unexpected persistence surface: monitored market activity and usage history remain on disk, which can leak information to other local users, backups, or forensic tooling and violates least-surprise for users expecting ephemeral behavior.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.