ClawHub
Back to skill

Security audit

Pdf Toolkit Free

Security checks across malware telemetry and agentic risk

Overview

The PDF skill mostly matches its stated purpose, but it includes under-scoped file, shell, and callback capabilities that deserve user review before installation.

Install only if you are comfortable with an agent using shell commands and writing output files for PDF work. Avoid using callback_url with sensitive documents unless you control the destination, and ask the agent to keep outputs in a specific folder and confirm before overwriting or deleting files.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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
Findings (3)

Context-Inappropriate Capability

Medium
Confidence
88% confidence
Finding
The skill accepts an arbitrary `callback_url` for a primarily local PDF-processing workflow, which unnecessarily expands the trust boundary to outbound network destinations. In an agent context, this can enable unintended data exfiltration, SSRF-like requests, or sending document-derived results to attacker-controlled endpoints without a tightly scoped business need.

Intent-Code Divergence

Medium
Confidence
77% confidence
Finding
The documentation inconsistently states that no API key is needed while also declaring an `LLM API` dependency as required. This mismatch can mislead users about external service use, causing them to unknowingly route document contents to third-party APIs or misconfigure the skill's trust and privacy assumptions.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The skill advertises file-writing and `exec` capabilities for document processing without prominent warnings about filesystem modification or shell-command execution. In an agent environment, this increases the chance of silent file changes, unsafe command construction, or overbroad execution against user files when triggered by natural-language requests.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.