ClawHub
Back to skill

Security audit

张洪Heo博客助手

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward helper for querying a public blog and its chatbot, with normal third-party content/link handling risk but no hidden install, persistence, credential use, or destructive behavior.

Install only if you are comfortable with the agent opening blog.zhheo.com, sending your chat/search prompt to that site, and returning the site's AI-generated text and links. Do not treat returned links or chatbot text as automatically trustworthy, and avoid sending private or sensitive information to the blog chatbot.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (4)

Ssd 3

Medium
Confidence
92% confidence
Finding
The skill instructs the agent to forward all AI chat content from the site to the user without filtering. That creates a data-handling risk because third-party chat output may include echoed user input, unexpected hidden content, links, or sensitive information that should be reviewed before relay.

Ssd 3

Medium
Confidence
93% confidence
Finding
The repeated directive to extract and relay all content from AI message bubbles normalizes unreviewed disclosure from an external system. Repetition increases the likelihood an agent will bypass normal safety checks and return unnecessary or unsafe content verbatim.

Ssd 3

Medium
Confidence
94% confidence
Finding
The DOM extraction instructions explicitly require collecting the full message body plus every reference link. This increases the chance of over-sharing untrusted or excessive content from the external chat UI, including malicious links or irrelevant embedded material.

Ssd 3

Medium
Confidence
92% confidence
Finding
The final note reinforces a behavior of forwarding entire external chat outputs without filtering, making the unsafe pattern operationally central to the skill. In context, the danger comes from unquestioned relay of third-party generated content rather than from the blog domain itself.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.