ClawHub
Back to skill

Security audit

Maxhub Threads

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed read-only Threads data lookup skill, but users should be mindful that it can query and chain public social profile and activity data through MaxHub.

Install only if you are comfortable sending Threads identifiers, search terms, URLs, and your MaxHub API key to MaxHub at aconfig.cn. Use explicit Threads-related prompts, avoid broad profiling or harassment, do not provide production account cookies or sessions, and minimize collection of personal social data.

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
93% confidence
Finding
The trigger list contains generic phrases such as “找人”, “搜索帖子”, and “看看评论”, which can match ordinary conversational language rather than a clear request to invoke this specific skill. In an agentic system, that can cause unintended recipe selection and downstream API calls, leading to accidental collection of social-media profile, post, or comment data beyond the user’s precise intent.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The trigger phrase set around ‘搜索用户 / 找人 / Threads用户’ is broad enough that normal conversational requests like ‘找人’ can unintentionally invoke this skill. In a skill that performs social-profile lookup and user analysis, overly broad activation increases the chance of unintended personal-data retrieval and cross-context tool use without clear user intent.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The documentation enables collection of third-party social-media profile, post, repost, and reply data and even encourages chained enumeration of identifiers, but it does not require user authorization, data-minimization, or privacy safeguards at the endpoint level. In a social-intelligence context this can facilitate disproportionate profiling or bulk collection of personal data, especially when usernames are resolved to stable user IDs and then expanded into activity history.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.