ClawHub

Knowledge Lib

Security checks across malware telemetry and agentic risk

Overview

This is a local markdown knowledge-base skill that persistently writes wiki files and logs, with behavior that matches its stated purpose.

Install this if you want an agent-managed local markdown wiki and are comfortable with it creating, updating, and logging files in the knowledge directory. Use explicit wiki commands, review KNOWLEDGE_BASE_DIR before running scripts, and avoid adding sensitive content unless you want it retained locally.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
84% confidence
Finding
The activation phrases include broad terms like "knowledge base," which can plausibly occur in ordinary conversation and accidentally trigger the skill. Because the skill has file-write permissions and performs persistent modifications, unintended activation could cause unrequested writes, summaries, or log entries in the user's workspace.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The quick-start guidance describes save/query actions without clearly warning that they create and modify files under a persistent workspace. Users may invoke the skill expecting ephemeral assistance, but the workflows append logs, write summaries, update concepts, and regenerate indexes, which can lead to unintended persistent data changes.

Missing User Warnings

Low
Confidence
86% confidence
Finding
The script writes a persistent entry to $KB_DIR/wiki/log.md every time it runs, but only informs the user after the write has already occurred. In an agent skill context, undisclosed persistent state changes can violate user expectations, create audit/privacy concerns, and gradually accumulate sensitive operational metadata such as dates and issue counts.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal