ClawHub
Back to skill

Security audit

Enhanced Prompt Suite

Security checks for vulnerabilities and agentic risk

Overview

This prompt-only skill does not run code, but it broadly guides the agent to produce high-stakes medical, legal, financial, and destructive-action plans without clear safeguards.

Install only if you want a very broad Chinese prompt-template suite. Treat its outputs as drafts or brainstorming, not professional medical, legal, financial, operational, or deletion instructions; require human review, authorization, backups, and applicable compliance checks before acting on high-stakes outputs.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (6)

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 4)May include surrounding context.

md
---
name: enhanced-prompt-suite
author: 王教成 Wang Jiaocheng (波动几何)
description: 增强提示词套件，为大语言模型安装操作系统、开发软件和应用软件的三层架构系统。①模式通用提示词套件（操作系统层，15个模式）：模式直用含连续场运算、身份涌现场、身份基元链；模式改进含10种创新元框架——第一性原理、逆向思维、辩证综合、随机性驱动、涌现生成、演化迭代、系统动力学、约束驱动、故事叙述、游戏化；以及模式迁移、模式构建。②领域专用提示词模板（开发软件层，8类模板）：通用模板可按任意领域生成

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill includes templates for disease symptoms, diagnosis, drugs, treatment, and psychological intervention, but provides no warning that outputs are informational only and not professional medical advice. This is dangerous because users may act on generated treatment or mental-health guidance directly, creating risk of physical harm, delayed care, or crisis escalation.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill provides broad legal, regulatory, rights-creation, contract, patent, licensing, and compliance-generation workflows without warning users that outputs may be jurisdiction-specific, incomplete, or legally invalid. Users could rely on generated legal documents or compliance conclusions, leading to rights loss, regulatory violations, contractual disputes, or unauthorized practice concerns.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill contains workflows for data clearing, archive destruction, and information注销/termination without safeguards, warnings, or verification requirements. Because these actions are destructive and often irreversible, generated instructions could cause permanent data loss, compliance breaches, or spoliation of records if followed improperly.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger list is extremely broad and contains many generic terms, making accidental or context-insensitive activation likely. In a skill that can generate medical, legal, security, and destructive-operation prompts, unintended activation materially increases the chance of unsafe workflow generation without deliberate user intent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill content and operational instructions are written entirely in Chinese, and the activation/usage description does not offer any user language choice or indicate that the skill is intentionally limited to a Chinese-speaking or region-specific context. Under the language/locale policy, a skill should not implicitly force a specific language without opt-in or a documented justification.

Content

No source excerpt is available for this finding.