ClawHub
Back to skill

Security audit

网页转飞书文档 Pro

Security checks across malware telemetry and agentic risk

Overview

This skill transparently converts user-provided webpages or files into Feishu documents, with expected cloud-write and credential use for that purpose.

Install only if you intend to let the agent fetch supplied links or files and create persistent Feishu documents. Use a least-privilege Feishu app, verify the target space or folder, and avoid submitting private files or sensitive URLs unless they are meant to be stored in Feishu.

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 (3)

Vague Triggers

Medium
Confidence
92% confidence
Finding
The trigger phrases are broad and generic for a write-capable skill that fetches arbitrary URLs/files and stores content into Feishu. This increases the chance of accidental activation in unrelated conversations, which could lead to unintended network access and document creation using external services.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The activation condition applies to essentially any URL or local file, without clear trust boundaries, file-type restrictions, or sensitivity checks. In a skill that can fetch content and persist it to Feishu, this can cause overbroad handling of sensitive local files or unintended external/ internal resources if invoked on loosely specified input.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill describes fetching arbitrary URLs/files and saving results to Feishu, but it does not clearly warn users up front that supplied content will be transmitted to external tools/services and persisted remotely. This weakens informed consent and can lead users to expose private URLs, documents, or regulated data without understanding the data flow.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.