ClawHub
Back to skill

Security audit

Niri IPC

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Niri desktop-control helper, but it gives an agent broad power to manipulate the active session and spawn shell commands without strong safety boundaries.

Install only if you want an agent to control your Niri desktop session. Prefer read-only queries by default, and require explicit user approval before raw IPC, closing windows, reloading config, moving workspaces, or using spawn/spawn-sh.

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 (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding
The skill documentation exposes capabilities to read environment state, invoke shell-accessible commands, and interact with IPC without declaring corresponding permissions or guardrails. In an agent setting, this creates a transparency and policy-bypass risk: callers may treat the skill as low-risk while it can execute impactful local actions such as spawning processes or reloading compositor configuration.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The documented actions include closing windows, moving windows/workspaces, reloading config, and spawning commands, but there is no warning that these operations can disrupt the user session, cause data loss in unsaved applications, or execute arbitrary programs. In this context, the absence of warnings and safety guidance makes it easier for an agent to perform destructive UI actions automatically and without user awareness.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.