ClawHub
Back to skill

Security audit

Alephnet节点

Security checks for vulnerabilities and agentic risk

Overview

This skill does not show clear malicious behavior, but it claims broad network, memory, social, wallet, and autonomous-agent powers while shipping only documentation and placeholder examples.

Review before installing. Treat this as an unverified specification/demo, not a working Alephnet client. Do not provide real API keys, wallet or staking credentials, private messages, group posting authority, broad filesystem access, or command execution approval until the publisher provides verifiable implementation, scoped permissions, explicit user confirmations, and clear data lifecycle controls.

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
  • YARA SignaturesMalware Match, Webshell Match, Cryptominer Match
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (4)

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 10)May include surrounding context.

md
�一等公民,系统封装语义场、分布式共识与 经济协议的复杂性,向上暴露高层认知"
summary_zh: "面向AI智能体的社会经济网络,提供语义计算、分布式记忆与一致性验证。面向AI智能体的完整社会经济网络。Agent作为一等公民,系统封装语义场、分布式共识与 经济协议的复杂性,向上暴露高层认知"
license: "MIT"
description: |-
  面向AI智能体的完整社会经济网络。Agent作为一等公民,系统封装语义场、分布式共识与
  经济协议的复杂性,向上暴露高层认知与社会动作。核心能力覆盖语义计算、分布式记忆场、
  社交图谱、消息系统、群组与信息流、一致性验证网络、智能体管理(SRIA)与积分经济.
  适用场景:多智能体协作、知识共识验证、分布式记忆存储、社群运营、自治学习.
  不适用于需要100%确定性的关键决策.
tags:
  - 研发工具
  - 工具
  - 效率
  - 写作

Intent-Code Divergence

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill claims substantial distributed-memory and network functionality, but the only concrete implementation shown is a placeholder command that prints a fixed string. This is dangerous because users or downstream agents may trust fabricated capabilities and make automation, security, or data-handling decisions based on nonexistent functionality, causing silent failure or unsafe fallback behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest presents the display name, summary, and summary_zh entirely in Chinese, and no alternate language or opt-in mechanism is provided elsewhere in the file. This can violate language/locale policy when a skill implicitly forces a specific language for all users rather than allowing user choice.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The documented JSON result format and execution log imply real processing of semantic/network tasks, yet the file provides no corresponding implementation and only placeholder behavior elsewhere. This can mislead agents into treating synthetic success responses as trustworthy outputs, which may propagate false state, incorrect audit records, or unsafe workflow continuation.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.