ClawHub
Back to skill

Security audit

Merge Pdf

Security checks across malware telemetry and agentic risk

Overview

This PDF merge skill is plausibly legitimate, but it asks for broad execution authority and sends user PDFs to an external service without enough scoping or data-handling detail.

Review before installing. Use this only for PDFs you are comfortable sending to Cross-Service-Solutions, and avoid sensitive legal, financial, medical, or confidential business documents unless you have separately verified the service's privacy and retention terms. The publisher should narrow the skill to PDF merging, remove or justify exec, and add an explicit confirmation step before any upload.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Context-Inappropriate Capability

Medium
Confidence
93% confidence
Finding
The skill is presented as a PDF merge utility, but its manifest grants `exec` capability without any clear, task-specific justification. Unnecessary command execution expands the attack surface substantially: if the agent later incorporates user-controlled filenames, paths, or instructions into shell commands, it can lead to arbitrary command execution or local system abuse.

Context-Inappropriate Capability

Medium
Confidence
86% confidence
Finding
The documentation advertises broad file conversion, content extraction, and command execution capabilities that go well beyond a narrowly scoped PDF merge tool. This kind of scope inflation encourages overbroad invocation and may cause agents to apply the skill to arbitrary file-processing or execution tasks, increasing the chance of unsafe handling of sensitive files or misuse of powerful tools.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The skill description gives overly broad guidance for generic file-processing, document conversion, and content extraction rather than limiting use to PDF merging. In agent ecosystems, ambiguous invocation scope is dangerous because it can cause the model to route unrelated or sensitive tasks through a skill that uploads files externally and exposes unnecessary capabilities.

Missing User Warnings

High
Confidence
97% confidence
Finding
The skill states that user PDFs are uploaded to an external service, but it does not present a clear upfront warning, consent flow, or data-handling disclosure before encouraging use. This is dangerous because PDFs often contain sensitive business, legal, or personal information, and silent third-party transfer can cause confidentiality breaches, compliance violations, and unintended data retention outside the user's environment.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.