ClawHub

Langchain Skill

Security checks across malware telemetry and agentic risk

Overview

This is a coherent LangChain assistant skill, but users should be careful with PDFs, local files, and external model providers.

Install only if you are comfortable with a LangChain-style assistant workflow. Do not provide confidential PDFs, local files, or secrets unless you know which model/search providers are configured and how uploaded content, memory, and any embeddings are stored or deleted.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (3)

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill advertises 'đọc file' capability but does not warn users that local files may be accessed or exposed through the agent workflow. In an agent context, undocumented file access increases the risk of unintentionally reading sensitive local data and then surfacing or transmitting it to external tools or model providers.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The documentation states that users can upload PDFs for RAG and question answering, but it does not warn how uploaded documents are stored, processed, retained, or shared. This can lead users to submit sensitive documents without understanding that their contents may be embedded, cached, or sent to third-party AI services.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The skill references external model providers and web search, but gives no warning that prompts, chat history, retrieved content, or tool inputs may be transmitted over the network. In a skill with memory, RAG, and tool calling, this omission is more dangerous because sensitive user data may be forwarded beyond the local environment without informed consent.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal