Security audit

ComPDF Convert Images To Documents

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent ComPDF image-conversion helper, but users should know it sends selected images to ComPDF and includes broader API reference material than the skill needs.

Install only if you are comfortable with selected images and derived document outputs being processed by ComPDF. Keep the API key file private, review requests before sending sensitive scans or receipts, and prefer pruning the bundled reference to image-conversion endpoints in a future release.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection

Findings (8)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 92% confidence
Finding: The skill instructs the agent to read a local API key file and send authenticated requests over the network, but it declares no permissions. This creates a capability mismatch that can bypass user expectations and safety controls, especially because the skill also references local file paths and external API usage.

Description-Behavior Mismatch

High

Confidence: 97% confidence
Finding: The reference file documents a huge set of capabilities far beyond the skill's declared purpose of converting images to documents, including PDF editing, encryption/decryption, AI extraction, and other document-processing actions. In an agent-skill context, overbroad documented capabilities can enable scope creep or accidental invocation of powerful actions the user did not intend, increasing the chance of unsafe data handling or misuse.

Context-Inappropriate Capability

High

Confidence: 98% confidence
Finding: Including PDF decryption in an image-conversion skill introduces a capability to remove document protections that is unrelated to the advertised purpose. This is dangerous because an LLM-driven agent may infer the capability is available and use it on protected files, producing unprotected outputs and undermining user expectations and document security controls.

Context-Inappropriate Capability

Medium

Confidence: 87% confidence
Finding: PDF encryption and permission-control features are not inherently malicious, but they are unrelated to the skill's narrow image-conversion scope. Their presence broadens the attack surface and can cause unintended file transformations or policy changes if the agent treats the reference as in-scope functionality.

Context-Inappropriate Capability

High

Confidence: 95% confidence
Finding: AI document parsing and extraction APIs materially expand the skill from simple format conversion into structured content extraction, including potentially sensitive field extraction from uploaded documents. In this skill context, that makes the mismatch more dangerous because users may believe they are only converting files while the available capabilities support deeper data extraction and broader external processing.

Context-Inappropriate Capability

Medium

Confidence: 84% confidence
Finding: PDF editing operations such as merge, split, delete, extract, insert, and rotate are outside the stated image-to-document conversion purpose. In an agent environment, unnecessary extra operations create opportunity for unintended document modification or destructive actions that the user did not request.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The document describes direct file upload and remote processing by ComPDF but does not clearly warn that user documents are being sent to an external third-party service. This is dangerous because users or downstream integrators may unintentionally transmit sensitive files off-platform without informed consent, creating privacy, compliance, and data-governance risk.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The decryption section documents removing password protection but lacks an explicit safety warning emphasizing that the output file will no longer be protected. In practice, this can lead to accidental creation and sharing of unencrypted sensitive documents, especially when embedded in an unrelated conversion skill.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.