ClawHub
Back to skill

Security audit

Search X

Security checks across malware telemetry and agentic risk

Overview

This is a coherent X/Twitter search skill that sends user-directed searches to xAI, with privacy and trigger clarity caveats but no evidence of hidden or harmful behavior.

Install this only if you are comfortable sending X/Twitter search terms, handle filters, and date filters to xAI using your configured API key. Avoid entering secrets, confidential project names, personal data, or sensitive investigative terms, and invoke it explicitly when you mean to search X/Twitter.

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 (7)

Missing User Warnings

Low
Confidence
90% confidence
Finding
The README explicitly states the skill uses xAI's Responses API and `x_search`, which means user-entered queries are sent to a third-party service. Without a clear privacy/data-handling warning, users may unknowingly transmit sensitive prompts, internal project names, or other confidential information off-platform. In this skill's context, external transmission is expected for functionality, which reduces severity, but the missing disclosure is still a real privacy and transparency issue.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The trigger phrase "find tweets" is broad and likely to match ordinary user requests, which can cause the skill to activate when the user did not intend to invoke this specific integration. In an agent ecosystem, overbroad activation can route user data and queries to an external API unexpectedly, creating privacy, consent, and command-selection risks.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The phrase "what's on x about" is ambiguous because "x" is heavily overloaded and may refer to a variable, a placeholder, the letter, or the social platform. This ambiguity increases the chance of accidental invocation and unintended transmission of prompts to the external X/xAI-backed search flow.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill transmits the user's search query, date range, and optional handle filters to xAI's external API, but the user-facing interface does not give an explicit privacy warning at execution time. This can expose sensitive search terms, targeted account lists, or investigative intent to a third party, which matters more in security, legal, or enterprise contexts where queries may contain confidential information.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
86% confidence
Finding
The trigger "search x" begins with the generic built-in command "search," creating a shadowing risk where the skill may intercept requests meant for the platform's standard search behavior. This can alter execution flow, reduce predictability, and send user queries to an unintended external service.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
86% confidence
Finding
The trigger "search twitter" similarly conflicts with a generic built-in "search" command and may hijack natural-language requests that should resolve to core search functionality. Because this skill sends requests to an external API, accidental interception has privacy and integrity implications beyond simple UX confusion.

Shadow Command Trigger

Medium
Category
Trigger Abuse
Confidence
87% confidence
Finding
The trigger "find tweets" overlaps with the built-in "find" command and can capture ordinary user intent in a way that redirects execution to this external-search skill. In context, this is more dangerous than a pure naming issue because the skill reaches outside the platform and may expose user queries or produce unexpected results from a different backend.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.