ClawHub
Back to skill

Security audit

Yuque 语雀知识库管理

Security checks across malware telemetry and agentic risk

Overview

This Yuque skill fits its stated purpose, but it stores an API token locally and can delete remote documents without clear safeguards.

Review before installing. Use a limited-scope Yuque token if possible, prefer an environment variable or protected secret storage over plaintext config.json, and require your agent to show the exact repo/document target and get explicit confirmation before any delete operation. Avoid using this for highly sensitive documents unless you are comfortable with content being written locally and sent to Yuque.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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
Findings (4)

Vague Triggers

Medium
Confidence
79% confidence
Finding
The invocation description covers general document search, read, create, edit, and note-management tasks, which could cause the skill to be selected for broad document workflows even when the user did not specifically intend Yuque. That increases the chance of unnecessary external API use, accidental data disclosure to Yuque, or writing content to the wrong destination.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The documentation exposes a delete capability for documents without any warning, confirmation step, or safeguard around destructive actions. In an agent setting, this raises the risk of accidental or overly broad deletion if the model misinterprets a request or the user phrasing is ambiguous.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The setup flow tells the agent to accept a user's API token, write it into config.json, and verify it, but provides no warning about credential sensitivity, storage location, file permissions, or avoiding logging the token. This can lead to long-lived secrets being stored insecurely in the skill directory, exposed through workspace sharing, backups, or accidental commits.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The script performs an irreversible DELETE operation directly from command-line arguments with no visible confirmation, dry-run mode, or safeguard against accidental invocation. In the context of a Yuque knowledge-base management skill, this is more dangerous because the tool is specifically designed to manage and modify user documents, so a mistaken parameter, automation error, or prompt-induced misuse could delete important content immediately.

VirusTotal

58/58 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.