Back to skill

Security audit

小红书爆款标题生成器 XHS Title Pro

Security checks for vulnerabilities and agentic risk

Overview

This is a Chinese Xiaohongshu title-writing helper with some overbroad routing and memory-style personalization notes, but no evidence of malicious code, exfiltration, hidden execution, or destructive behavior.

Install only if you want a Chinese Xiaohongshu-focused title assistant. Review the host agent's memory settings before use, and avoid granting file read/write access unless you specifically want it to work with local draft files.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

YARA rule 'agent_skill_mcp_tool_poisoning_metadata': MCP/tool metadata poisoning indicators in tool schemas or skill manifests [agent_skills]

High
Category
YARA Match
Confidence
80% confidence
Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · SKILL.md (reported line 3)May include surrounding context.

md
---
name: xhs-title-pro
description: >
  使用场景:
  - 用户说"帮我起标题"、"标题怎么写"
  - 用户说"小红书标题"、"爆款标题"
  - 用户粘贴笔记内容说"帮我起个好标题"
  - 用户说"笔记没人看"、"怎么涨粉"
  - 用户说"这个内容怎么取标题"

  不适用:
  - 公众号文章标题(用公众号标题生成器)
  - 抖音视频标题(用抖音标题生成器)
  - 简历优化(用resume-pro)
  - 纯闲聊
tools:
  - Read
  - Write
---

# 📌 小红书爆款标题生成器 (XHS Title Pro)

## Identity & Memory

你是**小红书爆款标题专家*

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

This markdown file contains user-facing skill instructions and examples exclusively in Chinese, which can amount to a language-policy issue when no opt-in or language-selection mechanism is provided. The document does not state that the skill is limited to Chinese-speaking users or a China-specific compliance context.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The activation scope is broad enough to match generic requests like 'help me write a title' or 'how to get more views,' which increases the chance the skill is invoked outside its intended Xiaohongshu-specific context. Over-broad routing can cause unintended data exposure to the skill and lead users into behavior the agent did not explicitly intend, especially when combined with memory instructions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill instructs the agent to remember user domain, audience, and style preferences across interactions without any user-facing notice or consent flow. This creates undisclosed retention of profiling data, which can violate privacy expectations and increase the blast radius if conversation history or memory is later accessed improperly.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The natural-language instruction to retain user-provided information across interactions is a direct memory directive. Without strict bounds, this can cause unnecessary storage of personal preferences and inferred profile data, increasing privacy risk and potentially affecting future responses in ways the user did not authorize.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The learning and memory sections describe ongoing retention of user feedback, effective formulas, and behavioral preferences, implying continuous profile-building without transparency. This is dangerous because it expands from simple task assistance into silent longitudinal tracking, which can be misused for inference about user interests and behavior over time.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill explicitly says it will learn from user feedback and remember high-performing combinations for future use, which operationalizes persistent retention beyond the immediate task. In a consumer-facing assistant, that creates a real risk of covert personalization and accumulation of behavioral data without clear boundaries.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The persona and tone instructions require use of youth-oriented Chinese internet slang and platform-native phrasing, which effectively forces a specific language/locale style. There is no indication that the user can choose another language or opt into this locale-specific mode.

Content

No source excerpt is available for this finding.