ClawHub
Back to skill

Security audit

Chinese DOCX Typesetting Baseline

Security checks for vulnerabilities and agentic risk

Overview

This skill provides document-formatting rules, with no evidence of hidden access, data theft, or unsafe automation.

Review shared-baseline edits because they affect future documents. Correct the extra quote in the version field and ensure the referenced conversion tools are available before use.

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
  • Rogue AgentSelf-Modification, Session Persistence
  • 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
Findings (2)

Self-Modification

High
Category
Rogue Agent
Confidence
85% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · SKILL.md (reported line 6)May include surrounding context.

md
line
version: "1.0.3""
description: "Locked design-token baseline for Chinese Word/DOCX documents and WeChat-native (公众号) article layouts: body 16px, section head 18px, title 23px, deep-blue #1a3a5c headings, no HTML tables/cards/backgrounds. Use when typesetting a Chinese DOCX or WeChat article, when html-to-docx output loses styles or fails on Windows, or when successive documents must stay visually consistent. Trigger keywords: 中文 docx 排版, Word 中文文档美化, html to docx 样式丢失, 16 18 23 px 基线, 深蓝标题, 公众号排版, 微信文章排版, WeChat article typesetting, 字号基线, font size baseline, Windows html-to-docx 失败, DOCX 样式. Baseline changes must update this skill first."
description_zh: "中文 Word 文档与公众号原生排版的锁定设计基线：正文 16px / 小标题 18px / 标题 23px、深蓝 #1a3a5c 标题色、禁 HTML 表格卡片背景、Windows html-to-docx 稳定转换路径。保证跨文档风格一致；基线变更必须先改本 skill。"
read_when: "Typesetting a Chinese DOCX or WeChat-native (公众号) article, aligning font/color tokens across documents, or when html-to-docx conversion loses styles or fails on Windows."
not_for:

Self-Modification

High
Category
Rogue Agent
Confidence
85% confidence
Finding

Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.

Content

Scanner excerpt · SKILL.md (reported line 43)May include surrounding context.

md
- Never flatten text to images; the DOCX must stay fully editable.
- WeChat-native layouts: no HTML tables, cards, background colors, or popup grids — vertical info bars only.
- Baseline changes (sizes, colors) must update this skill first, then the documents. No per-task deviation.
- Headings stay anchored to topic content (who / scene / problem), never filename-style titles.
- Line and paragraph spacing must be declared explicitly in the template; never inherit converter defaults, which differ between WPS and Word and silently produce inconsistent documents.

Static analysis

No suspicious patterns detected.