ClawHub

aaa

Security checks across malware telemetry and agentic risk

Overview

This is a transparent Web of Science to Feishu workflow, but users should confirm scope and destination before letting it access accounts or write records.

Install this only if you intend to use your own WoS/SZU and Feishu access for literature collection. Before writeback, confirm the topic, target count, screening rules, lark-cli account, Feishu Base link, subtable, field changes, and append-versus-overwrite behavior; do not store passwords or verification codes.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger condition is overly broad because it activates on any mention of WOS/Web of Science, even when the user may only be asking a conceptual question rather than requesting browser automation, institutional login, or Feishu writeback. In this skill, broad auto-activation is more dangerous because the workflow includes external site access, possible credential-handling prompts, and data writes to a Feishu Base, so mistaken invocation could lead to unnecessary collection actions or unintended modifications.

Vague Triggers

Medium
Confidence
83% confidence
Finding
The default prompt invokes a broad workflow that searches Web of Science, screens papers, and writes results into Feishu Base without any stated approval, scope, or trigger constraints. This can cause overbroad or unintended data operations if the skill is invoked in ambiguous contexts, especially because it combines external retrieval, judgment, and downstream write actions.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal