ClawHub
Back to skill

Security audit

Taizi Knowledge Base

Security checks across malware telemetry and agentic risk

Overview

This is a local personal knowledge-base skill that stores user-selected information, with privacy cautions but no evidence of hidden or malicious behavior.

Install only if you want a local memory and knowledge-base workflow. Review the referenced vector_kb.py script before relying on it, use explicit save commands, and avoid storing private conversations or files unless you intend them to remain searchable later.

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 (3)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The trigger phrase set includes broad, natural-language phrases such as “帮我记住”, which can easily appear in ordinary conversation and cause the skill to activate unintentionally. Because the skill performs persistent writes to local knowledge, ontology, and notes stores, accidental activation can store sensitive or irrelevant user content without clear consent.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The activation section defines triggers but provides no clear boundaries, disambiguation rules, or exclusions for normal conversational use. In this skill’s context, that is risky because activation leads directly into automated classification and storage workflows, increasing the chance of unintended persistence of user data.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill documents persistent storage locations for vector data, ontology data, and notes, but does not warn users that their content may be written to long-lived local storage. This omission is security-relevant because users may unknowingly disclose sensitive information that becomes indexed, searchable, and retained across sessions.

VirusTotal

51/51 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.