Back to skill

Security audit

barkpush

Security checks across malware telemetry and agentic risk

Overview

The skill appears to be a legitimate Bark notification helper, but users should know it sends notification content to Bark and can keep local history with message text and device keys.

Install only if you trust the configured Bark service and are comfortable with push content, URLs, recipient device keys, and push parameters being sent to it. Avoid putting secrets in notifications, protect ~/.bark-push/config.json and ~/.bark-push/history.json, consider disabling enable_update if you do not need update/delete history, and review broadcast/delete commands before using them.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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)

Lp3

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding
The skill declares no required permissions, yet its documented behavior clearly involves reading configuration, writing history/state, using environment data, and sending data over the network to the Bark service. This creates a transparency and consent gap: users and platforms may underestimate the skill's capabilities, increasing the risk of unintended data exposure or unsafe deployment.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill description explains how to push messages but does not clearly warn that message content, user aliases/device keys, and related push parameters will be transmitted to an external Bark service. Without explicit disclosure, users may send sensitive text or identifiers assuming the operation is local, causing confidentiality and privacy risks.

Missing User Warnings

Low
Confidence
82% confidence
Finding
The skill supports updating and deleting previously pushed notifications, but the documentation does not prominently warn that these operations can modify or remove existing user-visible messages. In multi-user or shared administrative contexts, this can lead to accidental tampering, confusion, or loss of important notification history.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
When enable_update is on, the service stores full push metadata including device keys, message body/markdown, user aliases, error details, and Bark API response data in history. Persisting sensitive content and identifiers without minimization, redaction, or visible consent increases the blast radius of local compromise and may expose private notification data unnecessarily.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
This code persists full notification history to disk, including device keys, user aliases, message bodies, parameters, error messages, and upstream response data, without any minimization, masking, retention protection, or access controls visible in this component. If the history file is readable by other local users, included in backups, or exposed through logs/support bundles, it can leak sensitive identifiers, notification content, and operational details that may enable privacy violations or abuse of Bark device keys.

Missing User Warnings

Medium
Confidence
87% confidence
Finding
The documentation explicitly demonstrates high-impact operations such as broadcasting to all users, updating, and deleting pushed messages, but provides no caution about accidental mass notification, destructive deletion, or the need for authorization and confirmation. In a push-notification skill, these examples can normalize risky usage and increase the chance of misuse or operator error, especially if copied directly into automation.

Missing User Warnings

Medium
Confidence
85% confidence
Finding
The usage guide shows sending arbitrary text, links, images, icons, and copyable content to an external Bark push service, but does not warn users that message bodies and referenced resources may be transmitted to or fetched through remote infrastructure. This creates privacy and data-handling risk because operators may unknowingly send sensitive content or cause clients to load external tracking resources.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.