ClawHub
Back to skill

Security audit

Raspberry Pi Health Check / 树莓派健康检查

Security checks for vulnerabilities and agentic risk

Overview

This skill is a clearly scoped Raspberry Pi health-check helper that discloses its optional JishuBuddy installation and keeps device actions read-only unless the user separately approves setup steps.

Before installing, confirm you want JishuBuddy installed globally on the operator machine, review the displayed npm version and global prefix, and disable telemetry with JISHUBUDDY_TELEMETRY_DISABLED=true before launch if desired. Use only trusted SSH targets and review any proposed SSH device configuration diff before approval.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding
The skill is presented primarily as a read-only Raspberry Pi health-check, but it also directs the agent to perform npm registry queries and optionally install JishuBuddy globally on the operator machine. That behavior expands the trust boundary and system impact beyond passive diagnostics, creating risk of unintended package installation, network access, and host modification if users or agents follow the workflow without clearly separating diagnostic steps from setup steps.

Ae1

High
Category
analysis-evasion
Content
`<skill-directory>` to the absolute directory containing this `SKILL.md`,
Confidence
100% confidence
Finding
Referenced artifact was not completely inspected

Static analysis

No suspicious patterns detected.