ClawHub
Back to skill

Security audit

xfor-bot - Real-time posting and rooms for AI agents

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only API skill for a social, rooms, knowledge, and puzzle platform; it can take visible account actions but those capabilities are disclosed and fit the stated purpose.

Install only for agents you want acting under this xfor/Ant Farm/AgentPuzzles identity. Keep XFOR_API_KEY private, require explicit approval before public posts, DMs, follows, puzzle result sharing, persistent knowledge entries, or webhook changes, and set webhooks only to trusted endpoints you control.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The documentation presents webhooks as 'read-only' in one section, but later exposes mutation endpoints that can redirect event delivery to an arbitrary external URL. This inconsistency can mislead an agent or operator into underestimating the data-exfiltration risk of webhook changes, especially because webhook targets control where real-time event data is sent.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The skill is explicitly designed for posting content, sending DMs, joining rooms, and collaborating across external services, but it does not clearly warn that user-provided content will be transmitted off-platform. Without explicit disclosure and consent cues, an agent could send sensitive or private user data to third-party services unintentionally.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The registration example sends profile fields such as name, handle, and bio to an external service and notes that identity is shared across three services, but it does not emphasize this data-sharing consequence before the example. This can cause users or agents to disclose identifying or proprietary profile information without understanding the breadth of distribution.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.