ClawHub
Back to skill

Security audit

一个智能的业务需求转研发文档工具。AI 自主分析项目代码库，理解业务需求，参考实际代码，生成可直接执行的研发文档。支持任意技术栈，无需配置。

Security checks for vulnerabilities and agentic risk

Overview

This skill is a documented requirements-writing helper that reads a project and stores a local project summary, with no evidence of hidden execution or data exfiltration.

Before installing, understand that the skill may read broad portions of a project and create a local `.ai-memory/project-profile.md` summary containing architecture, paths, stack details, and code patterns. Review that file in private repositories and add `.ai-memory/` to ignore rules if you do not want it committed.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (8)

Hidden Instructions

High
Category
Prompt Injection
Content
## 3. 页面设计
```html
<!-- 工具栏新增按钮 -->
<div class="toolbar">
  <el-button type="primary" @click="handleBatchExport">
    <el-icon><Download /></el-icon>
Confidence
70% confidence
Finding
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Natural-Language Policy Violations

Medium
Confidence
95% confidence
Finding
The skill example and instructional content are entirely in Chinese, including headings, prompts, and outputs, with no indication that users can choose another language. Under the stated policy, forcing a specific language without opt-in is a natural-language policy concern unless the locale limitation is explicitly justified.

Natural-Language Policy Violations

Medium
Confidence
88% confidence
Finding
The skill content is written as Chinese-only and presents outputs in Chinese without offering a user language choice. This is primarily a usability and operational safety issue: users who cannot fully read the generated requirements or warnings may misunderstand persisted-data behavior, implementation guidance, or security-relevant caveats.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill explicitly says it will store project analysis results under `.ai-memory/`, but it does not clearly warn users up front that repository-derived information will be persisted to disk. In a codebase context, that memory file may capture sensitive architectural details, internal paths, API patterns, or business logic summaries, creating unintended data retention and disclosure risk.

Missing User Warnings

Low
Confidence
90% confidence
Finding
This markdown example shows the skill saving data to `~/projects/skye-oms/.ai-memory/project-profile.md`, which affects user files. The surrounding description presents the behavior as part of execution but does not include any warning or disclosure that the skill may write to the local filesystem.

Missing User Warnings

Low
Confidence
83% confidence
Finding
The example requirement says the system will automatically send notifications to procurement staff, and later references possible channels including email or enterprise messaging. The markdown does not include any warning that the skill may generate or recommend behavior involving outbound notifications, which can affect privacy or system operations.

Natural-Language Policy Violations

Low
Confidence
78% confidence
Finding
The README presents the skill description and usage examples entirely in Chinese and does not state whether other languages are supported or let users choose their preferred language. Under the policy, forcing or implicitly requiring a specific language without opt-in can be a natural-language policy issue.

Missing User Warnings

Low
Confidence
94% confidence
Finding
The README explicitly states that the agent will persist a project profile to `.ai-memory/project-profile.md`, which may include repository structure, stack details, and implementation patterns. While this is not inherently malicious, storing repository-derived metadata locally without warning, scoping guidance, or sensitivity limits can expose internal architecture details and create unintended data retention risks, especially in private or regulated codebases.

Static analysis

No suspicious patterns detected.