ClawHub
Back to skill

Security audit

加密交易机器人

Security checks for vulnerabilities and agentic risk

Overview

This skill is a crypto trading assistant that can guide live trades and use exchange secrets, but it does not clearly limit or warn about those high-risk actions.

Review carefully before installing. Use this only for paper trading or read-only analysis unless you intentionally enable live trading. Do not paste real exchange secrets into chat or logs; use environment variables or a secrets manager, create least-privilege API keys, disable withdrawals, and require explicit confirmation for any real-money order.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (5)

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The skill asserts that 'risky code was removed' and security was enhanced, yet it still declares the powerful `exec` capability. That mismatch can mislead users and downstream agents into overtrusting the skill while leaving a broad execution primitive available, increasing the chance of unsafe command execution or abuse.

Context-Inappropriate Capability

Medium
Confidence
88% confidence
Finding
The skill exposes generic command execution for a crypto trading assistant without clearly constraining why shell access is needed. In an agent setting, this can expand the attack surface from data analysis into arbitrary local command execution, enabling file access, credential exposure, or system modification if the agent follows unsafe prompts.

Vague Triggers

Medium
Confidence
84% confidence
Finding
The activation text says to use the skill whenever the user needs crypto trading bot functionality, which is overly broad for a capability that includes live trading and command execution. Broad triggers raise the likelihood that the skill is invoked in marginal or ambiguous contexts, causing unintended trading-related actions or exposure to higher-risk tools.

Missing User Warnings

High
Confidence
95% confidence
Finding
The skill describes live trading execution and automated order placement without prominent warnings that these are real-money, irreversible actions. In the crypto context, this materially increases danger because mistaken prompts, misconfiguration, or prompt injection could trigger financial loss very quickly on volatile markets.

Missing User Warnings

High
Confidence
97% confidence
Finding
The quick-start and configuration examples instruct users to supply exchange API keys and secrets, but they do not include strong warnings about secret handling or the risk of enabling live trading. This is especially dangerous in an agent environment because users may paste production credentials into logs, prompts, config files, or insecure storage, leading to account compromise and unauthorized trades.

Static analysis

No suspicious patterns detected.