ClawHub
Back to skill

Security audit

Notion命令行(专业版)

Security checks across malware telemetry and agentic risk

Overview

This Notion CLI skill is not clearly malicious, but it asks for broad command-driven access to Notion data and external destinations without enough scoping or user control.

Install only if you intend to let an agent run Notion CLI commands that can read, modify, move, delete, export, and sync Notion workspace data. Use least-privilege Notion tokens, avoid broad workspace keys, review every write/delete/export command before execution, and do not enable warehouse, webhook, S3, scheduling, or Redis features unless you have verified the destinations and retention settings yourself.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Description-Behavior Mismatch

Medium
Confidence
89% confidence
Finding
The skill is framed as a Notion CLI tool, but it also advertises outbound sync to databases, webhook delivery, and S3 export. This capability expansion increases data-exfiltration and unintended third-party transfer risk, especially because the document does not define clear trust boundaries, approval gates, or destination allowlists for these external integrations.

Intent-Code Divergence

Medium
Confidence
95% confidence
Finding
The security section claims command execution is restricted to a whitelist, but the skill content only shows generic exec scaffolding and no actual enforcement mechanism. In an agent context with exec capability enabled, this mismatch can lead users or downstream systems to overtrust the skill and permit arbitrary command execution or parameter injection paths.

Vague Triggers

Medium
Confidence
77% confidence
Finding
The activation condition is broad enough to trigger this skill for loosely related requests, increasing the chance that an agent will invoke a powerful exec-enabled tool outside its intended scope. Because the skill handles credentials, workspace management, and write operations, overbroad routing can cause unintended data access or modification.

Vague Triggers

Medium
Confidence
80% confidence
Finding
The startup timing language references generic file processing and document conversion rather than specific Notion operations, which can cause misrouting into an exec-capable skill for unrelated tasks. In a tool-enabled agent environment, vague startup guidance broadens the operational surface and raises the chance of unsafe invocation.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.