Pdf Anthropic

Security checks across malware telemetry and agentic risk

Overview

This is a coherent PDF-processing skill that works on local PDF files and creates derived files, with some user-control cautions but no evidence of hidden or malicious behavior.

Install this only if you want the agent to process local PDFs. Treat PDFs as sensitive: confirm which files are being read, write outputs to new filenames, keep backups before repair operations, and only decrypt or remove passwords from documents you are authorized to access.

SkillSpector

By NVIDIA

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 93% confidence
Finding: The skill clearly demonstrates file read and file write capabilities through examples that open local PDFs, create outputs, and save derived files, yet no permissions are declared. That creates a trust and enforcement gap: an agent may invoke the skill with filesystem access broader than the user expects, increasing the chance of unauthorized document access or accidental overwrite of sensitive files.

Vague Triggers

High

Confidence: 89% confidence
Finding: The activation rule is extremely broad: it instructs use of the skill whenever the user wants to do anything with PDFs or even merely mentions a .pdf file. Overbroad triggering can cause the agent to route many ordinary requests through a high-capability file-processing skill unnecessarily, increasing exposure to untrusted documents, unintended file access, and tool misuse.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal