ClawHub
Back to skill

Security audit

Linear同步(专业版)

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed Linear automation helper, but it gives agents broad authority to change or delete workspace data and includes confirmation-bypassing delete examples.

Install only if you want an agent to manage live Linear data. Before using destructive or GraphQL mutation commands, require the agent to show the exact target issue, document, project, or team change and get explicit approval; avoid `-y` delete examples unless you have confirmed recovery and permissions.

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 (4)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The skill declares activation scope using extremely broad scenario keywords rather than narrow trigger conditions, which can cause an agent to invoke this powerful skill in unrelated contexts. Because the skill exposes exec, GraphQL, project/team management, and destructive operations, overbroad matching materially increases the chance of unintended execution and downstream harmful actions.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The documentation includes document deletion commands without a prominent warning about data loss, required authorization, or the consequences of using confirmation-bypass flags. In an agent setting, examples often become executable guidance, so normalizing deletion commands can lead to accidental destructive actions against live workspace data.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill shows forced task deletion using the -y flag, which bypasses confirmation and encourages non-interactive destructive behavior. Even though the text later mentions soft delete, the example still increases the risk that an agent will delete tasks automatically or without adequate human review.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The example demonstrates retrieving an auth token and immediately placing it into an HTTP Authorization header for network transmission, without a nearby warning about secret handling, shell history, process inspection, logging, or least-privilege use. In an AI-agent or shared environment, this pattern can expose credentials through command traces or copied examples and makes secret exfiltration easier if the command target is modified.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.