Back to skill

Security audit

[Nyx Archive] Companion Agent

Security checks across malware telemetry and agentic risk

Overview

This is an instruction-only skill for setting up a companion agent, and its Telegram and shared-memory behavior is disclosed and aligned with that purpose, though users should handle the bot token and shared workspace carefully.

Install only if you want a companion agent that shares the workspace and can interact through Telegram. Store the Telegram bot token securely, avoid committing config with secrets, restrict who can message the bot, keep sensitive personal or operational details out of shared memory unless intentional, and review any recurring heartbeat or cron behavior before enabling it.

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill instructs users to place a Telegram bot token directly into configuration but does not warn that the token is a secret, should be stored securely, or must not be committed to shared files. In a shared workspace or source-controlled config, this can lead to credential leakage and unauthorized control of the bot account.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill encourages Telegram DM integration and shared session behavior without warning that prompts, memory content, and heartbeat messages may be transmitted to and stored by an external messaging service. Because the companion shares the same workspace and can access identity and memory files, users may unintentionally expose sensitive personal or operational data through Telegram.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.