Back to skill

Security audit

quant-strategy-interpreter

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed Chinese-language quant research and backtesting skill with educational scripts, not a trading or credential-access tool.

Install this only if you want a Chinese quant-research aid. Treat all strategies and backtests as educational, verify any external data MCP you connect, and do not use outputs as investment advice or live-trading instructions without independent review.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (34)

Tp4

High
Category
MCP Tool Poisoning
Confidence
92% confidence
Finding

声明描述的是一个较完整的量化策略研究助手,核心能力包括文献解读、中文转述、回测模板生成、参数说明,以及基于真实市场数据回测。实际代码则是单一的 backtest_guardrails.py 演示脚本,使用模拟数据对简单移动均线参数进行训练/测试划分与样本外衰减检查,主要用于说明过拟合风险控制。虽然它属于量化回测相关范畴,并且确实是 Python/pandas 代码,但其主要目的和能力范围明显窄于且不同于声明,尤其缺失文献解读与真实行情数据接入这两项关键承诺,因此应判定为描述与行为不匹配。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

声明的核心能力是“量化/期货文献策略解读 + 回测模板生成 + 真实行情回测”。而代码的核心行为只是一个简化的权益组合构建示例:单因子动量选股 + 逆波动率配权 + Sharpe 计算。虽然它属于广义量化范畴,和“策略/回测”主题有弱相关,但与声明中的主要用途存在实质偏差:一是资产类别不符,代码做的是 equity portfolio,而不是期货策略;二是没有文献解读、中文说明、参数解释、出处汇总等“助手型”输出;三是没有接入任何 MCP 或真实行情数据,反而使用随机生成价格序列;四是并非通用模板或多策略框架,而是单一固定演示脚本。因此描述不能准确代表该代码的实际行为。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

声明描述的是一个面向量化/期货策略研究与回测的综合助手,重点能力包括文献解读、策略复现、参数说明以及基于真实行情的数据回测。实际代码却只是一个非常简化的微观结构示例函数 fill_prob(queue_pos, lam),用于说明队列位置越靠后成交概率越低,并在命令行打印若干样例。这属于交易微观结构/高频成交建模的教学演示,与所声明的主要用途相比偏离明显。虽然都处于广义量化交易领域,但该代码既没有策略回测,也没有文献解析管线,更没有数据获取或期货策略实现,因此描述不能准确代表代码实际行为。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

声明强调的是“策略解读 + 回测模板 + 参数说明 + 文献出处 + 真实行情回测”的综合型量化/期货助手,触发词也集中在双均线、海龟、均值回归、套利等具体策略方向。但实际代码仅是一个教学性质的统计演示脚本,围绕 Dacorogna 的高频时间缩放规律和波动聚集现象进行模拟与打印结果。它没有交易信号、持仓逻辑、绩效评估、参数配置说明、中文文献解读流程,也没有任何真实数据查询或 MCP 调用。因此其主要目的与声明的核心用途存在明显偏离,属于实质性描述不匹配。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

声明的核心功能是“文献策略解读 + 回测模板 + 参数说明 + 真实行情回测”,偏向研究/教学型量化策略助手;而代码的核心功能是高频执行层面的 VWAP/TWAP 拆单示例,仅演示订单执行切片和与日内成交量轮廓的拟合误差。这属于交易执行算法,不是文献策略复现或期货策略回测。虽然都属于广义量化交易领域,但主用途存在实质差异,且代码没有实现声明中的关键能力(回测、真实数据查询、策略解读)。因此应判定为描述与实际行为不匹配。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

声明描述的是一个面向量化/期货研究的综合助手,核心能力包括文献策略解读、中文转述、回测模板生成、参数说明、文献出处整理以及真实行情回测。实际代码只是一个独立的 Kelly 公式演示脚本,输入为胜率、盈亏比或收益序列,输出 Kelly 和半 Kelly 仓位比例。它属于资金管理中的单一计算工具,既没有解析英文文献,也没有实现双均线、海龟、均值回归、布林带、套利等任一触发词相关策略,更没有接入外部数据资源进行回测。因此其主要用途与声明严重不符,应判定为描述与行为不匹配。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

声明描述的是一个面向量化交易和期货策略研究的助手,核心能力应包括策略思想解读、回测模板生成、参数说明、文献出处整理,以及结合真实行情做回测。实际代码则是单一的期权 Black-Scholes 定价与 Greeks 演示脚本,只进行数学计算和示例输出。虽然期权也属于金融量化领域的相关主题,但其功能与“策略解读 + 回测 + 数据调用”的主目标差异显著,且代码完全没有实现声明中的关键能力,因此属于明显不匹配。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

声明的核心用途是‘量化交易与期货策略解读 + 回测’,重点应在策略思想转译、回测模板、参数说明、文献出处以及真实数据回测。实际代码则聚焦于金融风险管理中的波动率与尾部风险度量,且只是一个自包含演示脚本:生成正态随机收益,计算 EWMA 波动率和历史模拟 VaR/ES,输出演示结果。这与交易策略、期货策略、双均线/海龟/套利等触发场景没有直接对应关系,也没有任何 MCP 数据调用、文献解读或回测功能。因此其主要功能与声明存在明显且实质性的偏差。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding

声明描述的是一个面向量化/期货策略研究的综合助手,重点包括文献解读、策略复现、参数解释和真实行情回测。而实际代码是一个独立的教学型脚本,只计算风险平价中的逆波动权重,并使用随机数据演示。它没有任何与自然语言解读、英文论文处理、期货策略说明、回测流程或外部数据调用相关的实现。虽然“风险平价”可勉强归于量化策略范畴,但代码的实际主用途与声明的综合助手能力相差较大,因此应判定为描述与行为不匹配。

Content

No source excerpt is available for this finding.

Tp4

High
Category
MCP Tool Poisoning
Confidence
94% confidence
Finding

声明描述的是一个面向多种量化/期货策略的综合助手,核心能力包括文献策略解读、回测模板生成以及基于真实行情的数据调用。实际代码却只是一个独立的 Python 脚本,用标准差估算年化波动并计算目标波动对应的仓位权重,属于单一策略组件/风险控制示例。它没有任何自然语言解读能力的体现,也没有回测流程、参数体系说明、文献出处管理,且数据来源是本地随机模拟而非真实行情或 MCP。虽然代码仍处于量化交易领域,和声明主题存在弱相关,但主功能与承诺能力明显不一致,因此应判定为描述与行为不匹配。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The file explicitly states the processing method is a Chinese methodology interpretation and further requires Chinese rewriting. This is a natural-language locale constraint with no opt-in or alternative language option, which can violate language-choice policy.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The file states that all subsequent strategy chapters will '只做中文解读' (Chinese-only interpretation). Because this is a natural-language instruction in a markdown file and no user language choice or opt-in is offered nearby, it may violate language/locale policy requirements.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The document states '处理方法:中文方法论解读' and '只做中文重写', which imposes a specific language for the skill's behavior. Under the policy, forcing a language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly justified or optional.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The file explicitly states the handling method is a Chinese methodology interpretation and Chinese rewrite, which imposes a specific language on the skill's output. Under the policy rules, language constraints should either be optional, user-selected, or clearly justified as region-specific; no such opt-in or justification is provided here.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The file provides concrete trading and short-selling strategy templates, parameter guidance, and references to real-market data backtesting without any explicit risk disclosure or warning about potential financial loss. In the context of a skill designed to interpret and operationalize quant strategies, this omission increases the chance that users treat the content as actionable advice and deploy risky strategies without understanding leverage, drawdowns, transaction costs, or model risk.

Content

No source excerpt is available for this finding.

Ae4

Medium
Category
analysis-evasion
Confidence
80% confidence
Finding

Suspicious Unicode normalization or mixed-script content

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The document presents all instructions, explanations, and usage guidance exclusively in Chinese. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale constraint is explicitly justified, which is not stated here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The module docstring and user-facing output are written in Chinese, including the main descriptive text and warnings, with no indication that users can choose another language. This creates a language/locale policy concern because the skill effectively requires Chinese comprehension without documented opt-in or justification for a region-specific audience.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The module docstring presents the skill description, usage, and behavior entirely in Chinese, and there is no indication that users can opt into another language. Under the policy, forcing a specific language without user choice is a natural-language policy violation unless the locale limitation is explicitly justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

All printed status and result messages are emitted only in Chinese, which enforces a single language during execution. Because the file does not provide a language option or explain a region-specific requirement, this violates the language/locale policy.

Content

No source excerpt is available for this finding.

Ae4

Medium
Category
analysis-evasion
Confidence
80% confidence
Finding

Suspicious Unicode normalization or mixed-script content

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

This code includes natural-language documentation and usage text entirely in Chinese, which imposes a specific language/locale on users. The stated policy flags language/locale constraints unless the skill offers a choice or clearly documents a justified region-specific limitation, which is not present here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The print statements shown to end users are exclusively in Chinese, so the script enforces a specific language during execution. There is no indication that users can opt into this locale or switch to another language, and no documented justification for a Chinese-only runtime interface.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The file’s descriptive text and usage guidance are written entirely in Chinese, and the final printed status message is also Chinese. This imposes a specific language on users without any opt-in, fallback, or indication that the skill is intentionally region-specific, which matches the language/locale policy violation criteria.

Content

No source excerpt is available for this finding.

Ae4

Medium
Category
analysis-evasion
Confidence
80% confidence
Finding

Suspicious Unicode normalization or mixed-script content

Content

No source excerpt is available for this finding.