Security audit

加密协议博弈论分析

Security checks for vulnerabilities and agentic risk

Overview

The skill is mainly an analytical crypto game-theory guide, but it also claims broad command, file, and API capabilities with inconsistent disclosure.

Review this skill before installing if your agent can execute commands or access local files. Its analytical content is coherent, but the operational claims are inconsistent; only use it in a constrained environment and avoid providing API keys or allowing file writes, callbacks, or command execution unless the exact action is user-directed and necessary.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
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
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Intent-Code Divergence

Medium

Confidence: 92% confidence
Finding: The skill gives conflicting statements about API key requirements, first saying no extra API key is needed and later saying some features require platform API keys via environment variables. This can mislead users and agents about what external services may be accessed and what secrets may be expected, increasing the chance of unsafe secret handling or unintended capability use.

Intent-Code Divergence

High

Confidence: 95% confidence
Finding: The document claims it cannot fetch real-time on-chain data, yet elsewhere advertises API integration and real-time threat-intelligence subscription. This mismatch obscures the true data-access surface of the skill, making it harder for users to judge whether network access, telemetry, or external data flows may occur.

Intent-Code Divergence

High

Confidence: 97% confidence
Finding: The skill is described in one section as pure Markdown instruction-driven operation, but other sections claim command execution, file operations, and external API usage. This materially misrepresents the operational behavior of the skill and can cause an agent or user to grant more trust than warranted to a skill with execution-capable tooling.

Vague Triggers

Medium

Confidence: 84% confidence
Finding: The activation language is broad and ambiguous, effectively encouraging invocation whenever a user needs vaguely related crypto game-theory functionality. Overbroad triggers increase the chance the skill is auto-selected in situations beyond its safe or intended scope, especially given the listed exec and external-operation capabilities.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The markdown advertises file handling, API integration, and command execution without clear user-facing warnings about side effects or operational impact. In an agent context, that omission can lead users to assume the skill is purely analytical when it may perform actions affecting the local environment or external systems.

Static analysis

No suspicious patterns detected.