Security audit

TextIn xParse Document Parse

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed document-parsing skill for xparse/TextIn with expected privacy and credential-handling considerations, not hidden or destructive behavior.

Install only if you are comfortable sending selected documents or document URLs to the xparse/TextIn service. Avoid using it on confidential files without approval, store exported results carefully, and prefer safer password handling over typing real document passwords directly into command examples.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
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
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The skill description is extremely broad and instructs the agent to prefer this skill over many other document-handling approaches for parsing, reading, searching, summarizing, extraction, debugging, and multi-document workflows across local files and URLs. That broad scope increases the chance the skill is invoked for loosely related requests, which can cause unnecessary transmission of sensitive documents or metadata to the external xparse service and expand the attack surface for over-collection or unintended tool use.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The API reference documents parsing local documents and document URLs through an external service and explicitly describes returning extracted text, images, tables, metadata, and Base64 image data, but it provides no privacy or data-handling warning. In this skill context, users may reasonably supply sensitive business or personal documents, so omission of an explicit disclosure can cause unintended exfiltration of confidential content to a third-party API.

Missing User Warnings

Low

Confidence: 84% confidence
Finding: The `--output` option is described only as saving results to a file, without warning that the file may contain highly sensitive extracted document text, tables, OCR output, and image-derived data. This can lead to accidental persistence of confidential information in insecure locations, especially in shared workspaces or agent runtimes that log or retain artifacts.

Missing User Warnings

Medium

Confidence: 92% confidence
Finding: The example `--password secret123` encourages passing sensitive material directly on the command line, which can leak through shell history, process listings, logs, telemetry, and screenshots. In this skill's context, users are likely handling real confidential documents, so normalizing insecure secret handling in documentation increases the chance of credential or document-password exposure during routine use.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.