ClawHub
Back to skill

Security audit

萃书

Security checks for vulnerabilities and agentic risk

Overview

The skill’s book-summary and infographic workflow is coherent, but it instructs the agent to automatically send a generated file through a local POPO transfer command without confirmation or clear recipient/path controls.

Review before installing. The core book extraction and infographic generation behavior is understandable, but the automatic POPO file transfer should be constrained or changed to show the generated image first and require explicit user approval before sending, with recipient and file path limited to the current request’s generated image.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Context-Inappropriate Capability

Medium
Confidence
95% confidence
Finding
The skill instructs the agent to invoke a local Node script to push a generated file directly to a recipient, which extends beyond normal content generation into external file delivery. This creates a capability boundary violation: if the receiver or file path is influenced by context or future changes, the skill could exfiltrate data or send unsolicited files without an explicit user approval step.

Missing User Warnings

High
Confidence
97% confidence
Finding
The skill requires automatically sending the generated image by file transfer immediately after generation, without any warning, consent, or preview step. In this context, the danger is increased because the skill includes an outbound transmission mechanism unrelated to minimal summarization, enabling unsolicited delivery and reducing the user’s opportunity to catch incorrect recipients, unsafe content, or unexpected attachments.

Static analysis

No suspicious patterns detected.