ClawHub

Deep Research

Security checks across malware telemetry and agentic risk

Overview

This deep-research skill is coherent, but it asks for broad local research access, an external service, subagents, and cross-conversation memory without clear user controls.

Install only if you are comfortable using an external research service with an API key, subagent delegation, local file/context analysis, and persistent cross-conversation memory. Before using it with private, regulated, or business-sensitive material, confirm retention, deletion, storage, and opt-out controls, and restrict which files and sources the agent may use.

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 (2)

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill explicitly advertises cross-conversation memory persistence but does not warn users what data may be retained, for how long, who can access it, or how sensitive information should be handled. In a research agent that aggregates files, search results, and iterative findings, this increases the chance that confidential prompts, proprietary documents, or personal data are retained unexpectedly and later exposed or misused.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The workflow explicitly instructs the agent to persist key findings and decisions for reuse across future conversations, which creates a cross-session data retention channel. Although the file includes broad authorization language, it does not provide a clear, specific warning about what data is stored, how long it is retained, where it is stored, or how users can control or disable retention, creating privacy and confidentiality risk.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal