AI 学术论文润色

Security checks across malware telemetry and agentic risk

Overview

The skill is coherent for academic polishing, but it asks users to send manuscript text and payment credentials to a plain-HTTP service without adequate privacy or security disclosure.

Review this carefully before installing or using it. Avoid sending confidential, unpublished, regulated, or proprietary manuscript text, and do not transmit payment credentials unless the service provides HTTPS, clear payment-token handling, and privacy/retention terms.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Missing User Warnings

Medium

Confidence: 97% confidence
Finding: The skill instructs users to send academic paper text to a remote endpoint over plain HTTP, which exposes potentially unpublished research, proprietary data, or personal information to interception or tampering in transit. The documentation also omits any privacy, retention, jurisdiction, or security notice, increasing the risk that users disclose sensitive manuscript content without informed consent.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The payment flow has users return payment credentials in headers and references an HTTP gateway, creating risk of credential exposure, replay, interception, or misuse if transmitted insecurely. The skill provides no warning or guidance on handling payment-related data, so users may unknowingly expose sensitive financial tokens to an untrusted or improperly secured service.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal