ClawHub
Back to skill

Security audit

PDF压缩工具（免费版）

Security checks across malware telemetry and agentic risk

Overview

This PDF compression skill does what it says, but it uploads local PDFs and an API key to a configurable external endpoint without enough privacy and endpoint-safety disclosure.

Review the API provider and endpoint before use. Only upload PDFs you are comfortable sending to that service, avoid sensitive documents unless the service is trusted, keep the API key out of logs and shared shell history, and prefer a fixed trusted HTTPS endpoint rather than arbitrary URLs.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Intent-Code Divergence

Medium
Confidence
98% confidence
Finding
The skill explicitly claims SSRF protection, but the validator only rejects a few localhost-style hostnames and allows many dangerous cases such as private RFC1918 ranges, link-local addresses, IPv6 loopback, DNS rebinding targets, and internal hostnames. In this skill context, the API endpoint is taken from an environment variable and then contacted with local file contents, so a misconfigured or attacker-influenced URL could cause sensitive PDF data or credentials to be sent to unintended internal or malicious destinations.

Missing User Warnings

High
Confidence
97% confidence
Finding
The skill is designed to upload user PDFs to an external API and later return a download link, yet it provides no meaningful warning about confidentiality, third-party processing, retention, jurisdiction, or link-sharing risks. Because PDFs often contain sensitive business or personal data, omitting this disclosure can lead users or agents to exfiltrate protected content to an untrusted external service without informed consent.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The documentation tells users to configure API keys via environment variables but omits basic credential-safety guidance such as avoiding logs, shell history leakage, shared environments, or accidental inclusion in code and screenshots. In an agent/exec-enabled skill, weak credential-handling guidance increases the chance of secret exposure during setup or troubleshooting.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.